Upload folder using huggingface_hub

README.md

@@ -0,0 +1,202 @@
+---
+library_name: peft
+base_model: Qwen/Qwen-VL-Chat
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.10.0

adapter_config.json

@@ -0,0 +1,380 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen-VL-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "transformer.h.17.attn.c_proj",
+    "transformer.h.20.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.1.attn.in_proj",
+    "transformer.h.3.attn.c_attn",
+    "transformer.visual.transformer.resblocks.12.attn.in_proj",
+    "transformer.visual.transformer.resblocks.47.attn.in_proj",
+    "transformer.h.28.mlp.w2",
+    "transformer.h.6.mlp.w2",
+    "transformer.h.13.mlp.w1",
+    "transformer.visual.transformer.resblocks.39.attn.out_proj",
+    "transformer.h.2.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.attn.out_proj",
+    "transformer.visual.transformer.resblocks.0.attn.out_proj",
+    "transformer.h.4.attn.c_proj",
+    "transformer.h.22.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.12.attn.out_proj",
+    "transformer.visual.transformer.resblocks.10.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.43.attn.in_proj",
+    "transformer.visual.transformer.resblocks.0.attn.in_proj",
+    "transformer.visual.transformer.resblocks.26.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.11.mlp.c_proj",
+    "transformer.h.0.attn.c_attn",
+    "transformer.h.19.mlp.w2",
+    "transformer.visual.transformer.resblocks.37.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.40.mlp.c_proj",
+    "transformer.h.31.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.mlp.c_fc",
+    "transformer.h.18.mlp.w1",
+    "transformer.h.23.mlp.w2",
+    "transformer.visual.transformer.resblocks.6.attn.out_proj",
+    "transformer.visual.transformer.resblocks.17.attn.in_proj",
+    "transformer.visual.transformer.resblocks.27.attn.out_proj",
+    "transformer.h.12.mlp.w2",
+    "transformer.h.23.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.29.attn.in_proj",
+    "transformer.h.10.mlp.w1",
+    "transformer.visual.transformer.resblocks.18.attn.out_proj",
+    "transformer.visual.transformer.resblocks.4.attn.out_proj",
+    "transformer.visual.transformer.resblocks.8.mlp.c_fc",
+    "transformer.h.9.mlp.w1",
+    "transformer.visual.transformer.resblocks.38.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.attn.in_proj",
+    "transformer.visual.transformer.resblocks.14.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.22.attn.in_proj",
+    "transformer.visual.transformer.resblocks.25.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.23.attn.out_proj",
+    "transformer.visual.transformer.resblocks.3.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.41.mlp.c_proj",
+    "transformer.h.24.attn.c_proj",
+    "transformer.visual.transformer.resblocks.7.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.38.mlp.c_fc",
+    "transformer.h.10.attn.c_attn",
+    "transformer.h.26.attn.c_attn",
+    "transformer.visual.transformer.resblocks.5.attn.in_proj",
+    "transformer.visual.transformer.resblocks.2.attn.out_proj",
+    "transformer.h.7.attn.c_proj",
+    "transformer.h.24.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.34.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.2.mlp.c_proj",
+    "transformer.h.12.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.14.attn.out_proj",
+    "transformer.h.18.attn.c_attn",
+    "transformer.h.23.attn.c_proj",
+    "transformer.h.27.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.26.mlp.c_proj",
+    "transformer.h.3.mlp.w1",
+    "transformer.h.2.mlp.w2",
+    "transformer.visual.transformer.resblocks.45.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.25.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.45.attn.out_proj",
+    "transformer.h.25.mlp.w1",
+    "transformer.visual.transformer.resblocks.15.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.24.attn.in_proj",
+    "transformer.h.1.attn.c_proj",
+    "transformer.h.1.attn.c_attn",
+    "transformer.visual.transformer.resblocks.4.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.9.mlp.c_fc",
+    "transformer.h.13.attn.c_attn",
+    "transformer.visual.transformer.resblocks.40.attn.out_proj",
+    "transformer.h.7.mlp.w2",
+    "transformer.h.9.attn.c_proj",
+    "transformer.h.15.attn.c_attn",
+    "transformer.visual.transformer.resblocks.0.mlp.c_fc",
+    "transformer.h.27.attn.c_attn",
+    "transformer.h.15.mlp.c_proj",
+    "transformer.h.21.mlp.w2",
+    "transformer.h.28.attn.c_proj",
+    "transformer.visual.transformer.resblocks.42.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.16.attn.out_proj",
+    "transformer.h.9.mlp.w2",
+    "transformer.visual.transformer.resblocks.9.attn.in_proj",
+    "transformer.visual.transformer.resblocks.28.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.mlp.c_proj",
+    "transformer.h.11.mlp.w1",
+    "transformer.visual.transformer.resblocks.18.attn.in_proj",
+    "transformer.h.10.attn.c_proj",
+    "transformer.visual.transformer.resblocks.42.mlp.c_fc",
+    "transformer.h.31.attn.c_attn",
+    "transformer.visual.transformer.resblocks.23.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.21.attn.in_proj",
+    "transformer.h.24.mlp.w1",
+    "transformer.visual.transformer.resblocks.35.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.7.mlp.c_proj",
+    "transformer.h.8.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.12.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.7.attn.out_proj",
+    "transformer.h.22.mlp.w2",
+    "transformer.h.29.mlp.w2",
+    "transformer.h.0.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.38.attn.in_proj",
+    "transformer.h.8.mlp.w1",
+    "transformer.h.0.mlp.w1",
+    "transformer.h.26.mlp.w2",
+    "transformer.h.25.attn.c_proj",
+    "transformer.h.27.mlp.w1",
+    "transformer.visual.transformer.resblocks.21.attn.out_proj",
+    "transformer.visual.transformer.resblocks.44.attn.in_proj",
+    "transformer.visual.transformer.resblocks.43.attn.out_proj",
+    "transformer.h.29.attn.c_attn",
+    "transformer.h.24.attn.c_attn",
+    "transformer.visual.transformer.resblocks.17.attn.out_proj",
+    "transformer.h.2.attn.c_proj",
+    "transformer.visual.transformer.resblocks.15.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.11.attn.in_proj",
+    "transformer.visual.transformer.resblocks.17.mlp.c_proj",
+    "transformer.h.11.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.41.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.5.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.18.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.24.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.attn.out_proj",
+    "transformer.h.1.mlp.w2",
+    "transformer.h.21.mlp.c_proj",
+    "transformer.h.23.attn.c_attn",
+    "transformer.visual.transformer.resblocks.34.attn.out_proj",
+    "transformer.h.14.attn.c_attn",
+    "transformer.visual.transformer.resblocks.2.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.31.attn.out_proj",
+    "transformer.visual.transformer.resblocks.30.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.11.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.31.attn.in_proj",
+    "transformer.visual.transformer.resblocks.39.mlp.c_proj",
+    "transformer.h.9.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.20.attn.out_proj",
+    "transformer.h.18.mlp.c_proj",
+    "transformer.h.19.mlp.w1",
+    "transformer.h.9.attn.c_attn",
+    "transformer.visual.transformer.resblocks.36.attn.out_proj",
+    "transformer.visual.transformer.resblocks.7.attn.in_proj",
+    "transformer.visual.transformer.resblocks.30.attn.in_proj",
+    "transformer.visual.transformer.resblocks.47.attn.out_proj",
+    "transformer.visual.transformer.resblocks.0.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.15.attn.in_proj",
+    "transformer.visual.transformer.resblocks.29.attn.out_proj",
+    "transformer.visual.transformer.resblocks.41.attn.in_proj",
+    "transformer.visual.transformer.resblocks.4.attn.in_proj",
+    "transformer.h.25.attn.c_attn",
+    "transformer.visual.transformer.resblocks.12.mlp.c_proj",
+    "transformer.h.16.mlp.w1",
+    "transformer.h.28.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.27.attn.in_proj",
+    "transformer.visual.transformer.resblocks.13.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.attn.in_proj",
+    "transformer.visual.transformer.resblocks.45.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.46.mlp.c_proj",
+    "transformer.h.30.mlp.w1",
+    "transformer.visual.transformer.resblocks.43.mlp.c_fc",
+    "transformer.h.15.mlp.w1",
+    "transformer.h.16.attn.c_proj",
+    "transformer.h.20.mlp.w1",
+    "transformer.visual.transformer.resblocks.21.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.10.mlp.c_proj",
+    "transformer.h.10.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.35.attn.in_proj",
+    "transformer.h.13.mlp.w2",
+    "transformer.visual.transformer.resblocks.8.attn.out_proj",
+    "transformer.visual.transformer.resblocks.20.mlp.c_proj",
+    "transformer.h.22.attn.c_proj",
+    "transformer.h.6.mlp.w1",
+    "transformer.h.18.mlp.w2",
+    "transformer.h.4.mlp.c_proj",
+    "transformer.h.3.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.42.attn.out_proj",
+    "transformer.visual.transformer.resblocks.36.attn.in_proj",
+    "transformer.visual.transformer.resblocks.17.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.43.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.37.attn.in_proj",
+    "transformer.visual.transformer.resblocks.1.attn.out_proj",
+    "transformer.visual.transformer.resblocks.22.mlp.c_fc",
+    "transformer.h.22.mlp.w1",
+    "transformer.visual.transformer.resblocks.44.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.37.attn.out_proj",
+    "transformer.visual.transformer.resblocks.34.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.29.mlp.c_fc",
+    "transformer.h.18.attn.c_proj",
+    "transformer.visual.transformer.resblocks.38.attn.out_proj",
+    "transformer.h.5.attn.c_attn",
+    "transformer.visual.transformer.resblocks.19.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.15.attn.out_proj",
+    "transformer.visual.transformer.resblocks.37.mlp.c_fc",
+    "transformer.h.5.attn.c_proj",
+    "transformer.h.7.attn.c_attn",
+    "transformer.visual.transformer.resblocks.28.attn.out_proj",
+    "transformer.visual.transformer.resblocks.31.mlp.c_proj",
+    "transformer.h.29.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.45.attn.in_proj",
+    "transformer.visual.transformer.resblocks.27.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.10.attn.out_proj",
+    "transformer.visual.transformer.resblocks.40.attn.in_proj",
+    "transformer.h.23.mlp.w1",
+    "transformer.visual.transformer.resblocks.28.attn.in_proj",
+    "transformer.h.12.attn.c_proj",
+    "transformer.h.16.mlp.w2",
+    "transformer.h.27.mlp.w2",
+    "transformer.visual.transformer.resblocks.22.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.9.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.47.mlp.c_proj",
+    "transformer.h.26.attn.c_proj",
+    "transformer.visual.transformer.resblocks.40.mlp.c_fc",
+    "transformer.h.8.mlp.w2",
+    "transformer.visual.transformer.resblocks.27.mlp.c_fc",
+    "transformer.h.17.mlp.w1",
+    "transformer.h.31.mlp.w2",
+    "transformer.visual.transformer.resblocks.11.attn.out_proj",
+    "transformer.h.28.mlp.w1",
+    "transformer.visual.transformer.resblocks.10.attn.in_proj",
+    "transformer.h.12.mlp.w1",
+    "transformer.h.30.mlp.w2",
+    "transformer.visual.transformer.resblocks.13.attn.in_proj",
+    "transformer.h.6.attn.c_attn",
+    "transformer.h.5.mlp.c_proj",
+    "transformer.h.6.mlp.c_proj",
+    "transformer.h.22.attn.c_attn",
+    "transformer.h.13.attn.c_proj",
+    "transformer.visual.transformer.resblocks.46.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.41.attn.out_proj",
+    "transformer.visual.transformer.resblocks.30.mlp.c_fc",
+    "transformer.h.17.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.5.attn.out_proj",
+    "transformer.h.4.mlp.w2",
+    "transformer.visual.transformer.resblocks.1.mlp.c_proj",
+    "transformer.h.11.mlp.w2",
+    "transformer.h.19.attn.c_attn",
+    "transformer.h.14.mlp.w1",
+    "transformer.visual.transformer.resblocks.44.attn.out_proj",
+    "transformer.visual.transformer.resblocks.14.mlp.c_fc",
+    "transformer.h.21.attn.c_attn",
+    "transformer.visual.transformer.resblocks.36.mlp.c_proj",
+    "transformer.h.2.mlp.w1",
+    "transformer.h.14.attn.c_proj",
+    "transformer.visual.transformer.resblocks.46.attn.in_proj",
+    "transformer.h.6.attn.c_proj",
+    "transformer.h.0.mlp.w2",
+    "transformer.h.5.mlp.w1",
+    "transformer.h.30.attn.c_proj",
+    "transformer.h.24.mlp.w2",
+    "transformer.h.0.attn.c_proj",
+    "transformer.visual.transformer.resblocks.4.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.22.attn.out_proj",
+    "transformer.h.10.mlp.w2",
+    "transformer.h.17.mlp.w2",
+    "transformer.visual.transformer.resblocks.23.attn.in_proj",
+    "transformer.visual.transformer.resblocks.36.mlp.c_fc",
+    "transformer.h.20.mlp.w2",
+    "transformer.visual.transformer.resblocks.9.attn.out_proj",
+    "transformer.h.29.mlp.w1",
+    "transformer.visual.transformer.resblocks.20.attn.in_proj",
+    "transformer.visual.transformer.resblocks.20.mlp.c_fc",
+    "transformer.h.15.attn.c_proj",
+    "transformer.h.3.mlp.w2",
+    "transformer.h.30.attn.c_attn",
+    "transformer.visual.transformer.resblocks.47.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.16.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.39.mlp.c_fc",
+    "transformer.h.20.attn.c_attn",
+    "transformer.h.19.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.46.attn.out_proj",
+    "transformer.visual.transformer.resblocks.29.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.19.attn.out_proj",
+    "transformer.visual.transformer.resblocks.26.attn.in_proj",
+    "transformer.visual.transformer.resblocks.16.mlp.c_fc",
+    "transformer.h.11.attn.c_proj",
+    "transformer.h.12.attn.c_attn",
+    "transformer.visual.conv1",
+    "transformer.visual.transformer.resblocks.35.attn.out_proj",
+    "transformer.h.25.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.14.attn.in_proj",
+    "transformer.h.26.mlp.w1",
+    "transformer.visual.transformer.resblocks.1.mlp.c_fc",
+    "transformer.h.7.mlp.c_proj",
+    "transformer.h.29.attn.c_proj",
+    "transformer.h.1.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.mlp.c_proj",
+    "transformer.h.14.mlp.c_proj",
+    "transformer.h.3.attn.c_proj",
+    "transformer.h.25.mlp.w2",
+    "transformer.h.20.attn.c_proj",
+    "transformer.h.16.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.attn.in_proj",
+    "transformer.h.17.attn.c_attn",
+    "transformer.h.14.mlp.w2",
+    "transformer.visual.transformer.resblocks.2.attn.in_proj",
+    "transformer.visual.transformer.resblocks.5.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.33.attn.out_proj",
+    "transformer.h.15.mlp.w2",
+    "transformer.h.4.attn.c_attn",
+    "transformer.h.31.mlp.w1",
+    "transformer.h.11.attn.c_attn",
+    "transformer.visual.transformer.resblocks.23.mlp.c_proj",
+    "transformer.h.7.mlp.w1",
+    "transformer.visual.transformer.resblocks.34.attn.in_proj",
+    "transformer.h.1.mlp.w1",
+    "transformer.visual.transformer.resblocks.28.mlp.c_fc",
+    "transformer.h.21.attn.c_proj",
+    "transformer.h.30.mlp.c_proj",
+    "transformer.h.21.mlp.w1",
+    "transformer.visual.transformer.resblocks.30.attn.out_proj",
+    "transformer.visual.transformer.resblocks.42.attn.in_proj",
+    "transformer.visual.transformer.resblocks.25.attn.out_proj",
+    "transformer.visual.transformer.resblocks.19.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.39.attn.in_proj",
+    "transformer.visual.transformer.resblocks.19.attn.in_proj",
+    "transformer.visual.transformer.resblocks.13.mlp.c_fc",
+    "transformer.h.13.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.25.attn.in_proj",
+    "transformer.visual.transformer.resblocks.31.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.24.attn.out_proj",
+    "transformer.visual.transformer.resblocks.24.mlp.c_fc",
+    "transformer.h.4.mlp.w1",
+    "transformer.h.8.attn.c_attn",
+    "transformer.visual.transformer.resblocks.21.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.44.mlp.c_proj",
+    "transformer.h.28.attn.c_attn",
+    "transformer.visual.transformer.resblocks.18.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.attn.in_proj",
+    "transformer.h.19.attn.c_proj",
+    "transformer.h.2.attn.c_attn",
+    "transformer.visual.transformer.resblocks.35.mlp.c_proj",
+    "transformer.h.26.mlp.c_proj",
+    "transformer.h.8.attn.c_proj",
+    "transformer.h.27.attn.c_proj",
+    "transformer.visual.transformer.resblocks.13.attn.out_proj",
+    "transformer.h.16.attn.c_attn",
+    "transformer.visual.transformer.resblocks.16.attn.in_proj",
+    "transformer.visual.transformer.resblocks.8.attn.in_proj",
+    "transformer.visual.transformer.resblocks.26.attn.out_proj",
+    "transformer.h.31.attn.c_proj",
+    "transformer.h.5.mlp.w2",
+    "transformer.visual.transformer.resblocks.8.mlp.c_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:29edeab874fb9c672969f9531b2a0473dab3b421c73f00c956d2cdc2b8e00b69
+size 469105640

checkpoint-1200/README.md

@@ -0,0 +1,202 @@
+---
+library_name: peft
+base_model: Qwen/Qwen-VL-Chat
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.10.0

checkpoint-1200/adapter_config.json

@@ -0,0 +1,380 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen-VL-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "transformer.h.17.attn.c_proj",
+    "transformer.h.20.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.1.attn.in_proj",
+    "transformer.h.3.attn.c_attn",
+    "transformer.visual.transformer.resblocks.12.attn.in_proj",
+    "transformer.visual.transformer.resblocks.47.attn.in_proj",
+    "transformer.h.28.mlp.w2",
+    "transformer.h.6.mlp.w2",
+    "transformer.h.13.mlp.w1",
+    "transformer.visual.transformer.resblocks.39.attn.out_proj",
+    "transformer.h.2.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.attn.out_proj",
+    "transformer.visual.transformer.resblocks.0.attn.out_proj",
+    "transformer.h.4.attn.c_proj",
+    "transformer.h.22.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.12.attn.out_proj",
+    "transformer.visual.transformer.resblocks.10.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.43.attn.in_proj",
+    "transformer.visual.transformer.resblocks.0.attn.in_proj",
+    "transformer.visual.transformer.resblocks.26.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.11.mlp.c_proj",
+    "transformer.h.0.attn.c_attn",
+    "transformer.h.19.mlp.w2",
+    "transformer.visual.transformer.resblocks.37.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.40.mlp.c_proj",
+    "transformer.h.31.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.mlp.c_fc",
+    "transformer.h.18.mlp.w1",
+    "transformer.h.23.mlp.w2",
+    "transformer.visual.transformer.resblocks.6.attn.out_proj",
+    "transformer.visual.transformer.resblocks.17.attn.in_proj",
+    "transformer.visual.transformer.resblocks.27.attn.out_proj",
+    "transformer.h.12.mlp.w2",
+    "transformer.h.23.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.29.attn.in_proj",
+    "transformer.h.10.mlp.w1",
+    "transformer.visual.transformer.resblocks.18.attn.out_proj",
+    "transformer.visual.transformer.resblocks.4.attn.out_proj",
+    "transformer.visual.transformer.resblocks.8.mlp.c_fc",
+    "transformer.h.9.mlp.w1",
+    "transformer.visual.transformer.resblocks.38.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.attn.in_proj",
+    "transformer.visual.transformer.resblocks.14.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.22.attn.in_proj",
+    "transformer.visual.transformer.resblocks.25.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.23.attn.out_proj",
+    "transformer.visual.transformer.resblocks.3.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.41.mlp.c_proj",
+    "transformer.h.24.attn.c_proj",
+    "transformer.visual.transformer.resblocks.7.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.38.mlp.c_fc",
+    "transformer.h.10.attn.c_attn",
+    "transformer.h.26.attn.c_attn",
+    "transformer.visual.transformer.resblocks.5.attn.in_proj",
+    "transformer.visual.transformer.resblocks.2.attn.out_proj",
+    "transformer.h.7.attn.c_proj",
+    "transformer.h.24.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.34.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.2.mlp.c_proj",
+    "transformer.h.12.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.14.attn.out_proj",
+    "transformer.h.18.attn.c_attn",
+    "transformer.h.23.attn.c_proj",
+    "transformer.h.27.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.26.mlp.c_proj",
+    "transformer.h.3.mlp.w1",
+    "transformer.h.2.mlp.w2",
+    "transformer.visual.transformer.resblocks.45.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.25.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.45.attn.out_proj",
+    "transformer.h.25.mlp.w1",
+    "transformer.visual.transformer.resblocks.15.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.24.attn.in_proj",
+    "transformer.h.1.attn.c_proj",
+    "transformer.h.1.attn.c_attn",
+    "transformer.visual.transformer.resblocks.4.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.9.mlp.c_fc",
+    "transformer.h.13.attn.c_attn",
+    "transformer.visual.transformer.resblocks.40.attn.out_proj",
+    "transformer.h.7.mlp.w2",
+    "transformer.h.9.attn.c_proj",
+    "transformer.h.15.attn.c_attn",
+    "transformer.visual.transformer.resblocks.0.mlp.c_fc",
+    "transformer.h.27.attn.c_attn",
+    "transformer.h.15.mlp.c_proj",
+    "transformer.h.21.mlp.w2",
+    "transformer.h.28.attn.c_proj",
+    "transformer.visual.transformer.resblocks.42.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.16.attn.out_proj",
+    "transformer.h.9.mlp.w2",
+    "transformer.visual.transformer.resblocks.9.attn.in_proj",
+    "transformer.visual.transformer.resblocks.28.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.mlp.c_proj",
+    "transformer.h.11.mlp.w1",
+    "transformer.visual.transformer.resblocks.18.attn.in_proj",
+    "transformer.h.10.attn.c_proj",
+    "transformer.visual.transformer.resblocks.42.mlp.c_fc",
+    "transformer.h.31.attn.c_attn",
+    "transformer.visual.transformer.resblocks.23.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.21.attn.in_proj",
+    "transformer.h.24.mlp.w1",
+    "transformer.visual.transformer.resblocks.35.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.7.mlp.c_proj",
+    "transformer.h.8.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.12.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.7.attn.out_proj",
+    "transformer.h.22.mlp.w2",
+    "transformer.h.29.mlp.w2",
+    "transformer.h.0.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.38.attn.in_proj",
+    "transformer.h.8.mlp.w1",
+    "transformer.h.0.mlp.w1",
+    "transformer.h.26.mlp.w2",
+    "transformer.h.25.attn.c_proj",
+    "transformer.h.27.mlp.w1",
+    "transformer.visual.transformer.resblocks.21.attn.out_proj",
+    "transformer.visual.transformer.resblocks.44.attn.in_proj",
+    "transformer.visual.transformer.resblocks.43.attn.out_proj",
+    "transformer.h.29.attn.c_attn",
+    "transformer.h.24.attn.c_attn",
+    "transformer.visual.transformer.resblocks.17.attn.out_proj",
+    "transformer.h.2.attn.c_proj",
+    "transformer.visual.transformer.resblocks.15.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.11.attn.in_proj",
+    "transformer.visual.transformer.resblocks.17.mlp.c_proj",
+    "transformer.h.11.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.41.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.5.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.18.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.24.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.attn.out_proj",
+    "transformer.h.1.mlp.w2",
+    "transformer.h.21.mlp.c_proj",
+    "transformer.h.23.attn.c_attn",
+    "transformer.visual.transformer.resblocks.34.attn.out_proj",
+    "transformer.h.14.attn.c_attn",
+    "transformer.visual.transformer.resblocks.2.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.31.attn.out_proj",
+    "transformer.visual.transformer.resblocks.30.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.11.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.31.attn.in_proj",
+    "transformer.visual.transformer.resblocks.39.mlp.c_proj",
+    "transformer.h.9.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.20.attn.out_proj",
+    "transformer.h.18.mlp.c_proj",
+    "transformer.h.19.mlp.w1",
+    "transformer.h.9.attn.c_attn",
+    "transformer.visual.transformer.resblocks.36.attn.out_proj",
+    "transformer.visual.transformer.resblocks.7.attn.in_proj",
+    "transformer.visual.transformer.resblocks.30.attn.in_proj",
+    "transformer.visual.transformer.resblocks.47.attn.out_proj",
+    "transformer.visual.transformer.resblocks.0.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.15.attn.in_proj",
+    "transformer.visual.transformer.resblocks.29.attn.out_proj",
+    "transformer.visual.transformer.resblocks.41.attn.in_proj",
+    "transformer.visual.transformer.resblocks.4.attn.in_proj",
+    "transformer.h.25.attn.c_attn",
+    "transformer.visual.transformer.resblocks.12.mlp.c_proj",
+    "transformer.h.16.mlp.w1",
+    "transformer.h.28.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.27.attn.in_proj",
+    "transformer.visual.transformer.resblocks.13.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.attn.in_proj",
+    "transformer.visual.transformer.resblocks.45.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.46.mlp.c_proj",
+    "transformer.h.30.mlp.w1",
+    "transformer.visual.transformer.resblocks.43.mlp.c_fc",
+    "transformer.h.15.mlp.w1",
+    "transformer.h.16.attn.c_proj",
+    "transformer.h.20.mlp.w1",
+    "transformer.visual.transformer.resblocks.21.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.10.mlp.c_proj",
+    "transformer.h.10.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.35.attn.in_proj",
+    "transformer.h.13.mlp.w2",
+    "transformer.visual.transformer.resblocks.8.attn.out_proj",
+    "transformer.visual.transformer.resblocks.20.mlp.c_proj",
+    "transformer.h.22.attn.c_proj",
+    "transformer.h.6.mlp.w1",
+    "transformer.h.18.mlp.w2",
+    "transformer.h.4.mlp.c_proj",
+    "transformer.h.3.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.42.attn.out_proj",
+    "transformer.visual.transformer.resblocks.36.attn.in_proj",
+    "transformer.visual.transformer.resblocks.17.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.43.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.37.attn.in_proj",
+    "transformer.visual.transformer.resblocks.1.attn.out_proj",
+    "transformer.visual.transformer.resblocks.22.mlp.c_fc",
+    "transformer.h.22.mlp.w1",
+    "transformer.visual.transformer.resblocks.44.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.37.attn.out_proj",
+    "transformer.visual.transformer.resblocks.34.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.29.mlp.c_fc",
+    "transformer.h.18.attn.c_proj",
+    "transformer.visual.transformer.resblocks.38.attn.out_proj",
+    "transformer.h.5.attn.c_attn",
+    "transformer.visual.transformer.resblocks.19.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.15.attn.out_proj",
+    "transformer.visual.transformer.resblocks.37.mlp.c_fc",
+    "transformer.h.5.attn.c_proj",
+    "transformer.h.7.attn.c_attn",
+    "transformer.visual.transformer.resblocks.28.attn.out_proj",
+    "transformer.visual.transformer.resblocks.31.mlp.c_proj",
+    "transformer.h.29.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.45.attn.in_proj",
+    "transformer.visual.transformer.resblocks.27.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.10.attn.out_proj",
+    "transformer.visual.transformer.resblocks.40.attn.in_proj",
+    "transformer.h.23.mlp.w1",
+    "transformer.visual.transformer.resblocks.28.attn.in_proj",
+    "transformer.h.12.attn.c_proj",
+    "transformer.h.16.mlp.w2",
+    "transformer.h.27.mlp.w2",
+    "transformer.visual.transformer.resblocks.22.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.9.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.47.mlp.c_proj",
+    "transformer.h.26.attn.c_proj",
+    "transformer.visual.transformer.resblocks.40.mlp.c_fc",
+    "transformer.h.8.mlp.w2",
+    "transformer.visual.transformer.resblocks.27.mlp.c_fc",
+    "transformer.h.17.mlp.w1",
+    "transformer.h.31.mlp.w2",
+    "transformer.visual.transformer.resblocks.11.attn.out_proj",
+    "transformer.h.28.mlp.w1",
+    "transformer.visual.transformer.resblocks.10.attn.in_proj",
+    "transformer.h.12.mlp.w1",
+    "transformer.h.30.mlp.w2",
+    "transformer.visual.transformer.resblocks.13.attn.in_proj",
+    "transformer.h.6.attn.c_attn",
+    "transformer.h.5.mlp.c_proj",
+    "transformer.h.6.mlp.c_proj",
+    "transformer.h.22.attn.c_attn",
+    "transformer.h.13.attn.c_proj",
+    "transformer.visual.transformer.resblocks.46.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.41.attn.out_proj",
+    "transformer.visual.transformer.resblocks.30.mlp.c_fc",
+    "transformer.h.17.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.5.attn.out_proj",
+    "transformer.h.4.mlp.w2",
+    "transformer.visual.transformer.resblocks.1.mlp.c_proj",
+    "transformer.h.11.mlp.w2",
+    "transformer.h.19.attn.c_attn",
+    "transformer.h.14.mlp.w1",
+    "transformer.visual.transformer.resblocks.44.attn.out_proj",
+    "transformer.visual.transformer.resblocks.14.mlp.c_fc",
+    "transformer.h.21.attn.c_attn",
+    "transformer.visual.transformer.resblocks.36.mlp.c_proj",
+    "transformer.h.2.mlp.w1",
+    "transformer.h.14.attn.c_proj",
+    "transformer.visual.transformer.resblocks.46.attn.in_proj",
+    "transformer.h.6.attn.c_proj",
+    "transformer.h.0.mlp.w2",
+    "transformer.h.5.mlp.w1",
+    "transformer.h.30.attn.c_proj",
+    "transformer.h.24.mlp.w2",
+    "transformer.h.0.attn.c_proj",
+    "transformer.visual.transformer.resblocks.4.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.22.attn.out_proj",
+    "transformer.h.10.mlp.w2",
+    "transformer.h.17.mlp.w2",
+    "transformer.visual.transformer.resblocks.23.attn.in_proj",
+    "transformer.visual.transformer.resblocks.36.mlp.c_fc",
+    "transformer.h.20.mlp.w2",
+    "transformer.visual.transformer.resblocks.9.attn.out_proj",
+    "transformer.h.29.mlp.w1",
+    "transformer.visual.transformer.resblocks.20.attn.in_proj",
+    "transformer.visual.transformer.resblocks.20.mlp.c_fc",
+    "transformer.h.15.attn.c_proj",
+    "transformer.h.3.mlp.w2",
+    "transformer.h.30.attn.c_attn",
+    "transformer.visual.transformer.resblocks.47.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.16.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.39.mlp.c_fc",
+    "transformer.h.20.attn.c_attn",
+    "transformer.h.19.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.46.attn.out_proj",
+    "transformer.visual.transformer.resblocks.29.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.19.attn.out_proj",
+    "transformer.visual.transformer.resblocks.26.attn.in_proj",
+    "transformer.visual.transformer.resblocks.16.mlp.c_fc",
+    "transformer.h.11.attn.c_proj",
+    "transformer.h.12.attn.c_attn",
+    "transformer.visual.conv1",
+    "transformer.visual.transformer.resblocks.35.attn.out_proj",
+    "transformer.h.25.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.14.attn.in_proj",
+    "transformer.h.26.mlp.w1",
+    "transformer.visual.transformer.resblocks.1.mlp.c_fc",
+    "transformer.h.7.mlp.c_proj",
+    "transformer.h.29.attn.c_proj",
+    "transformer.h.1.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.mlp.c_proj",
+    "transformer.h.14.mlp.c_proj",
+    "transformer.h.3.attn.c_proj",
+    "transformer.h.25.mlp.w2",
+    "transformer.h.20.attn.c_proj",
+    "transformer.h.16.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.attn.in_proj",
+    "transformer.h.17.attn.c_attn",
+    "transformer.h.14.mlp.w2",
+    "transformer.visual.transformer.resblocks.2.attn.in_proj",
+    "transformer.visual.transformer.resblocks.5.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.33.attn.out_proj",
+    "transformer.h.15.mlp.w2",
+    "transformer.h.4.attn.c_attn",
+    "transformer.h.31.mlp.w1",
+    "transformer.h.11.attn.c_attn",
+    "transformer.visual.transformer.resblocks.23.mlp.c_proj",
+    "transformer.h.7.mlp.w1",
+    "transformer.visual.transformer.resblocks.34.attn.in_proj",
+    "transformer.h.1.mlp.w1",
+    "transformer.visual.transformer.resblocks.28.mlp.c_fc",
+    "transformer.h.21.attn.c_proj",
+    "transformer.h.30.mlp.c_proj",
+    "transformer.h.21.mlp.w1",
+    "transformer.visual.transformer.resblocks.30.attn.out_proj",
+    "transformer.visual.transformer.resblocks.42.attn.in_proj",
+    "transformer.visual.transformer.resblocks.25.attn.out_proj",
+    "transformer.visual.transformer.resblocks.19.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.39.attn.in_proj",
+    "transformer.visual.transformer.resblocks.19.attn.in_proj",
+    "transformer.visual.transformer.resblocks.13.mlp.c_fc",
+    "transformer.h.13.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.25.attn.in_proj",
+    "transformer.visual.transformer.resblocks.31.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.24.attn.out_proj",
+    "transformer.visual.transformer.resblocks.24.mlp.c_fc",
+    "transformer.h.4.mlp.w1",
+    "transformer.h.8.attn.c_attn",
+    "transformer.visual.transformer.resblocks.21.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.44.mlp.c_proj",
+    "transformer.h.28.attn.c_attn",
+    "transformer.visual.transformer.resblocks.18.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.attn.in_proj",
+    "transformer.h.19.attn.c_proj",
+    "transformer.h.2.attn.c_attn",
+    "transformer.visual.transformer.resblocks.35.mlp.c_proj",
+    "transformer.h.26.mlp.c_proj",
+    "transformer.h.8.attn.c_proj",
+    "transformer.h.27.attn.c_proj",
+    "transformer.visual.transformer.resblocks.13.attn.out_proj",
+    "transformer.h.16.attn.c_attn",
+    "transformer.visual.transformer.resblocks.16.attn.in_proj",
+    "transformer.visual.transformer.resblocks.8.attn.in_proj",
+    "transformer.visual.transformer.resblocks.26.attn.out_proj",
+    "transformer.h.31.attn.c_proj",
+    "transformer.h.5.mlp.w2",
+    "transformer.visual.transformer.resblocks.8.mlp.c_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-1200/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42041b221d86a9753fed065745dc7df697e59016b80d615003f1759f4ad6203d
+size 469105640

checkpoint-1200/latest

@@ -0,0 +1 @@
+global_step1200

checkpoint-1200/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:336248bbe8a4fee02df88588f7f7dc1b33253e35723db0c2b4226da31752a2d3
+size 14960

checkpoint-1200/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e29d9da4dc40c9b6b09b727757e4b19b448bc1dfeb00627e256a8e07f67e4da9
+size 14960

checkpoint-1200/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4ae0fe36d0ba0ead4de4005e87904f5e4d9dec09a80b2e4db5ec0c80a0ea346
+size 14960

checkpoint-1200/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9087f0ffb3a26097f004f56dccdbf08eec0c5cc75577bc9d741246ab7c60a229
+size 14960

checkpoint-1200/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80613ac1f3063b2760ed75829d953d81b43e14572849068a9f7570742ebc5962
+size 1064

checkpoint-1200/special_tokens_map.json

@@ -0,0 +1,3 @@
+{
+  "pad_token": "<|endoftext|>"
+}

checkpoint-1200/tokenizer_config.json

@@ -0,0 +1,14 @@
+{
+  "added_tokens_decoder": {},
+  "auto_map": {
+    "AutoTokenizer": [
+      "Qwen/Qwen-VL-Chat--tokenization_qwen.QWenTokenizer",
+      null
+    ]
+  },
+  "clean_up_tokenization_spaces": true,
+  "model_max_length": 1280,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "tokenizer_class": "QWenTokenizer"
+}

checkpoint-1200/trainer_state.json

@@ -0,0 +1,873 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.20191822311963656,
+  "eval_steps": 500,
+  "global_step": 1200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016826518593303045,
+      "grad_norm": 5.367858933563703,
+      "learning_rate": 4.9999999999999996e-06,
+      "loss": 0.9537,
+      "step": 10
+    },
+    {
+      "epoch": 0.003365303718660609,
+      "grad_norm": 9.386746384686745,
+      "learning_rate": 9.999999999999999e-06,
+      "loss": 0.943,
+      "step": 20
+    },
+    {
+      "epoch": 0.005047955577990914,
+      "grad_norm": 7.387362447577942,
+      "learning_rate": 1.5e-05,
+      "loss": 0.934,
+      "step": 30
+    },
+    {
+      "epoch": 0.006730607437321218,
+      "grad_norm": 6.9256319824932655,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 0.8376,
+      "step": 40
+    },
+    {
+      "epoch": 0.008413259296651522,
+      "grad_norm": 9.1148382590838,
+      "learning_rate": 2.5e-05,
+      "loss": 0.8484,
+      "step": 50
+    },
+    {
+      "epoch": 0.010095911155981827,
+      "grad_norm": 3.9989232759892426,
+      "learning_rate": 3e-05,
+      "loss": 0.8097,
+      "step": 60
+    },
+    {
+      "epoch": 0.011778563015312132,
+      "grad_norm": 3.892371218590039,
+      "learning_rate": 2.9999786123888308e-05,
+      "loss": 0.7811,
+      "step": 70
+    },
+    {
+      "epoch": 0.013461214874642436,
+      "grad_norm": 8.096662196282066,
+      "learning_rate": 2.9999144501652298e-05,
+      "loss": 0.7446,
+      "step": 80
+    },
+    {
+      "epoch": 0.01514386673397274,
+      "grad_norm": 1.5769306611206149,
+      "learning_rate": 2.9998075151588992e-05,
+      "loss": 0.7258,
+      "step": 90
+    },
+    {
+      "epoch": 0.016826518593303044,
+      "grad_norm": 8.47430485487969,
+      "learning_rate": 2.999657810419285e-05,
+      "loss": 0.7052,
+      "step": 100
+    },
+    {
+      "epoch": 0.01850917045263335,
+      "grad_norm": 2.363071299913598,
+      "learning_rate": 2.999465340215489e-05,
+      "loss": 0.7657,
+      "step": 110
+    },
+    {
+      "epoch": 0.020191822311963654,
+      "grad_norm": 1.9252385425154874,
+      "learning_rate": 2.999230110036149e-05,
+      "loss": 0.7329,
+      "step": 120
+    },
+    {
+      "epoch": 0.02187447417129396,
+      "grad_norm": 8.946028475031488,
+      "learning_rate": 2.99895212658928e-05,
+      "loss": 0.7304,
+      "step": 130
+    },
+    {
+      "epoch": 0.023557126030624265,
+      "grad_norm": 6.877609312630206,
+      "learning_rate": 2.9986313978020846e-05,
+      "loss": 0.7453,
+      "step": 140
+    },
+    {
+      "epoch": 0.02523977788995457,
+      "grad_norm": 2.5256324882367993,
+      "learning_rate": 2.9982679328207262e-05,
+      "loss": 0.7366,
+      "step": 150
+    },
+    {
+      "epoch": 0.02692242974928487,
+      "grad_norm": 2.709550398238738,
+      "learning_rate": 2.9978617420100692e-05,
+      "loss": 0.7258,
+      "step": 160
+    },
+    {
+      "epoch": 0.028605081608615177,
+      "grad_norm": 1.543550019689774,
+      "learning_rate": 2.9974128369533805e-05,
+      "loss": 0.7372,
+      "step": 170
+    },
+    {
+      "epoch": 0.03028773346794548,
+      "grad_norm": 3.3453966881155504,
+      "learning_rate": 2.9969212304520034e-05,
+      "loss": 0.743,
+      "step": 180
+    },
+    {
+      "epoch": 0.03197038532727579,
+      "grad_norm": 1.922001656181265,
+      "learning_rate": 2.9963869365249895e-05,
+      "loss": 0.7819,
+      "step": 190
+    },
+    {
+      "epoch": 0.03365303718660609,
+      "grad_norm": 2.0611188483400036,
+      "learning_rate": 2.995809970408699e-05,
+      "loss": 0.7155,
+      "step": 200
+    },
+    {
+      "epoch": 0.0353356890459364,
+      "grad_norm": 1.5313041833127259,
+      "learning_rate": 2.9951903485563685e-05,
+      "loss": 0.7322,
+      "step": 210
+    },
+    {
+      "epoch": 0.0370183409052667,
+      "grad_norm": 2.0124191694435085,
+      "learning_rate": 2.99452808863764e-05,
+      "loss": 0.6759,
+      "step": 220
+    },
+    {
+      "epoch": 0.03870099276459701,
+      "grad_norm": 3.182123324389477,
+      "learning_rate": 2.993823209538056e-05,
+      "loss": 0.6953,
+      "step": 230
+    },
+    {
+      "epoch": 0.04038364462392731,
+      "grad_norm": 1.6122782177661379,
+      "learning_rate": 2.9930757313585238e-05,
+      "loss": 0.6953,
+      "step": 240
+    },
+    {
+      "epoch": 0.04206629648325761,
+      "grad_norm": 2.2027482596695647,
+      "learning_rate": 2.9922856754147406e-05,
+      "loss": 0.7301,
+      "step": 250
+    },
+    {
+      "epoch": 0.04374894834258792,
+      "grad_norm": 2.6782477155989213,
+      "learning_rate": 2.9914530642365852e-05,
+      "loss": 0.6891,
+      "step": 260
+    },
+    {
+      "epoch": 0.04543160020191822,
+      "grad_norm": 1.9740401144541417,
+      "learning_rate": 2.990577921567476e-05,
+      "loss": 0.7231,
+      "step": 270
+    },
+    {
+      "epoch": 0.04711425206124853,
+      "grad_norm": 1.719874620968932,
+      "learning_rate": 2.989660272363696e-05,
+      "loss": 0.7505,
+      "step": 280
+    },
+    {
+      "epoch": 0.04879690392057883,
+      "grad_norm": 1.3138364164203409,
+      "learning_rate": 2.988700142793676e-05,
+      "loss": 0.7116,
+      "step": 290
+    },
+    {
+      "epoch": 0.05047955577990914,
+      "grad_norm": 5.853627389344256,
+      "learning_rate": 2.9876975602372536e-05,
+      "loss": 0.719,
+      "step": 300
+    },
+    {
+      "epoch": 0.05216220763923944,
+      "grad_norm": 2.347259437170711,
+      "learning_rate": 2.9866525532848906e-05,
+      "loss": 0.6803,
+      "step": 310
+    },
+    {
+      "epoch": 0.05384485949856974,
+      "grad_norm": 1.937679220955038,
+      "learning_rate": 2.9855651517368567e-05,
+      "loss": 0.7461,
+      "step": 320
+    },
+    {
+      "epoch": 0.05552751135790005,
+      "grad_norm": 1.6661300351569575,
+      "learning_rate": 2.9844353866023802e-05,
+      "loss": 0.7472,
+      "step": 330
+    },
+    {
+      "epoch": 0.05721016321723035,
+      "grad_norm": 2.357915869204484,
+      "learning_rate": 2.9832632900987642e-05,
+      "loss": 0.7148,
+      "step": 340
+    },
+    {
+      "epoch": 0.05889281507656066,
+      "grad_norm": 4.398815186243292,
+      "learning_rate": 2.982048895650468e-05,
+      "loss": 0.6992,
+      "step": 350
+    },
+    {
+      "epoch": 0.06057546693589096,
+      "grad_norm": 12.662682224480092,
+      "learning_rate": 2.9807922378881537e-05,
+      "loss": 0.7539,
+      "step": 360
+    },
+    {
+      "epoch": 0.06225811879522127,
+      "grad_norm": 0.8642696401357872,
+      "learning_rate": 2.979493352647697e-05,
+      "loss": 0.7212,
+      "step": 370
+    },
+    {
+      "epoch": 0.06394077065455157,
+      "grad_norm": 27.047937858232604,
+      "learning_rate": 2.9781522769691686e-05,
+      "loss": 0.722,
+      "step": 380
+    },
+    {
+      "epoch": 0.06562342251388188,
+      "grad_norm": 2.598805292448644,
+      "learning_rate": 2.9767690490957758e-05,
+      "loss": 0.7065,
+      "step": 390
+    },
+    {
+      "epoch": 0.06730607437321218,
+      "grad_norm": 1.2314762895092763,
+      "learning_rate": 2.9753437084727713e-05,
+      "loss": 0.7498,
+      "step": 400
+    },
+    {
+      "epoch": 0.06898872623254249,
+      "grad_norm": 1.6421909669790502,
+      "learning_rate": 2.9738762957463292e-05,
+      "loss": 0.6992,
+      "step": 410
+    },
+    {
+      "epoch": 0.0706713780918728,
+      "grad_norm": 2.023552968622588,
+      "learning_rate": 2.9723668527623877e-05,
+      "loss": 0.6943,
+      "step": 420
+    },
+    {
+      "epoch": 0.0723540299512031,
+      "grad_norm": 1.5172337910969138,
+      "learning_rate": 2.9708154225654526e-05,
+      "loss": 0.6987,
+      "step": 430
+    },
+    {
+      "epoch": 0.0740366818105334,
+      "grad_norm": 1.197852135730745,
+      "learning_rate": 2.9692220493973712e-05,
+      "loss": 0.7302,
+      "step": 440
+    },
+    {
+      "epoch": 0.0757193336698637,
+      "grad_norm": 2.4396443837967183,
+      "learning_rate": 2.9675867786960718e-05,
+      "loss": 0.7318,
+      "step": 450
+    },
+    {
+      "epoch": 0.07740198552919401,
+      "grad_norm": 1.4599851880563282,
+      "learning_rate": 2.9659096570942654e-05,
+      "loss": 0.6941,
+      "step": 460
+    },
+    {
+      "epoch": 0.07908463738852431,
+      "grad_norm": 1.117755825364562,
+      "learning_rate": 2.9641907324181194e-05,
+      "loss": 0.7399,
+      "step": 470
+    },
+    {
+      "epoch": 0.08076728924785462,
+      "grad_norm": 2.9235378164576242,
+      "learning_rate": 2.96243005368589e-05,
+      "loss": 0.7207,
+      "step": 480
+    },
+    {
+      "epoch": 0.08244994110718493,
+      "grad_norm": 7.308883163781362,
+      "learning_rate": 2.960627671106527e-05,
+      "loss": 0.682,
+      "step": 490
+    },
+    {
+      "epoch": 0.08413259296651522,
+      "grad_norm": 3.4394827932955234,
+      "learning_rate": 2.9587836360782405e-05,
+      "loss": 0.708,
+      "step": 500
+    },
+    {
+      "epoch": 0.08581524482584553,
+      "grad_norm": 3.2314529856927634,
+      "learning_rate": 2.9568980011870357e-05,
+      "loss": 0.7335,
+      "step": 510
+    },
+    {
+      "epoch": 0.08749789668517584,
+      "grad_norm": 1.825724533695325,
+      "learning_rate": 2.954970820205214e-05,
+      "loss": 0.6951,
+      "step": 520
+    },
+    {
+      "epoch": 0.08918054854450615,
+      "grad_norm": 3.3231741746640076,
+      "learning_rate": 2.9530021480898393e-05,
+      "loss": 0.7793,
+      "step": 530
+    },
+    {
+      "epoch": 0.09086320040383644,
+      "grad_norm": 1.3097651462571123,
+      "learning_rate": 2.9509920409811696e-05,
+      "loss": 0.7087,
+      "step": 540
+    },
+    {
+      "epoch": 0.09254585226316675,
+      "grad_norm": 6.685911471215255,
+      "learning_rate": 2.9489405562010565e-05,
+      "loss": 0.6906,
+      "step": 550
+    },
+    {
+      "epoch": 0.09422850412249706,
+      "grad_norm": 2.870746617513948,
+      "learning_rate": 2.9468477522513132e-05,
+      "loss": 0.7028,
+      "step": 560
+    },
+    {
+      "epoch": 0.09591115598182735,
+      "grad_norm": 1.782555352805469,
+      "learning_rate": 2.9447136888120408e-05,
+      "loss": 0.6901,
+      "step": 570
+    },
+    {
+      "epoch": 0.09759380784115766,
+      "grad_norm": 2.336519711000487,
+      "learning_rate": 2.9425384267399327e-05,
+      "loss": 0.7779,
+      "step": 580
+    },
+    {
+      "epoch": 0.09927645970048797,
+      "grad_norm": 8.935574410818228,
+      "learning_rate": 2.940322028066534e-05,
+      "loss": 0.7503,
+      "step": 590
+    },
+    {
+      "epoch": 0.10095911155981828,
+      "grad_norm": 2.754713786882031,
+      "learning_rate": 2.938064555996476e-05,
+      "loss": 0.7208,
+      "step": 600
+    },
+    {
+      "epoch": 0.10264176341914857,
+      "grad_norm": 1.5082503557652136,
+      "learning_rate": 2.9357660749056713e-05,
+      "loss": 0.7169,
+      "step": 610
+    },
+    {
+      "epoch": 0.10432441527847888,
+      "grad_norm": 9.04522194526273,
+      "learning_rate": 2.9334266503394803e-05,
+      "loss": 0.6927,
+      "step": 620
+    },
+    {
+      "epoch": 0.10600706713780919,
+      "grad_norm": 55.28278686388287,
+      "learning_rate": 2.9310463490108397e-05,
+      "loss": 0.7107,
+      "step": 630
+    },
+    {
+      "epoch": 0.10768971899713949,
+      "grad_norm": 3.721916069105249,
+      "learning_rate": 2.928625238798362e-05,
+      "loss": 0.6951,
+      "step": 640
+    },
+    {
+      "epoch": 0.1093723708564698,
+      "grad_norm": 2.5040797323750112,
+      "learning_rate": 2.9261633887443993e-05,
+      "loss": 0.6916,
+      "step": 650
+    },
+    {
+      "epoch": 0.1110550227158001,
+      "grad_norm": 3.5468924769840617,
+      "learning_rate": 2.9236608690530738e-05,
+      "loss": 0.7077,
+      "step": 660
+    },
+    {
+      "epoch": 0.11273767457513041,
+      "grad_norm": 3.0266819778200746,
+      "learning_rate": 2.921117751088276e-05,
+      "loss": 0.6952,
+      "step": 670
+    },
+    {
+      "epoch": 0.1144203264344607,
+      "grad_norm": 1.634743894298146,
+      "learning_rate": 2.91853410737163e-05,
+      "loss": 0.6936,
+      "step": 680
+    },
+    {
+      "epoch": 0.11610297829379101,
+      "grad_norm": 1.0925365801520501,
+      "learning_rate": 2.915910011580426e-05,
+      "loss": 0.7317,
+      "step": 690
+    },
+    {
+      "epoch": 0.11778563015312132,
+      "grad_norm": 1.6959112138540386,
+      "learning_rate": 2.9132455385455176e-05,
+      "loss": 0.6917,
+      "step": 700
+    },
+    {
+      "epoch": 0.11946828201245162,
+      "grad_norm": 1.9723433746891168,
+      "learning_rate": 2.9105407642491895e-05,
+      "loss": 0.7209,
+      "step": 710
+    },
+    {
+      "epoch": 0.12115093387178193,
+      "grad_norm": 2.1537215293733833,
+      "learning_rate": 2.907795765822989e-05,
+      "loss": 0.7488,
+      "step": 720
+    },
+    {
+      "epoch": 0.12283358573111224,
+      "grad_norm": 3.227101869737169,
+      "learning_rate": 2.9050106215455283e-05,
+      "loss": 0.7152,
+      "step": 730
+    },
+    {
+      "epoch": 0.12451623759044254,
+      "grad_norm": 2.7222358893572554,
+      "learning_rate": 2.9021854108402516e-05,
+      "loss": 0.708,
+      "step": 740
+    },
+    {
+      "epoch": 0.12619888944977284,
+      "grad_norm": 2.1054843767538136,
+      "learning_rate": 2.8993202142731693e-05,
+      "loss": 0.7251,
+      "step": 750
+    },
+    {
+      "epoch": 0.12788154130910315,
+      "grad_norm": 2.11845883419618,
+      "learning_rate": 2.8964151135505616e-05,
+      "loss": 0.7405,
+      "step": 760
+    },
+    {
+      "epoch": 0.12956419316843346,
+      "grad_norm": 13.171512404187755,
+      "learning_rate": 2.8934701915166477e-05,
+      "loss": 0.6844,
+      "step": 770
+    },
+    {
+      "epoch": 0.13124684502776376,
+      "grad_norm": 2.7633375632879127,
+      "learning_rate": 2.890485532151225e-05,
+      "loss": 0.6766,
+      "step": 780
+    },
+    {
+      "epoch": 0.13292949688709407,
+      "grad_norm": 1.8420785342693768,
+      "learning_rate": 2.887461220567271e-05,
+      "loss": 0.7037,
+      "step": 790
+    },
+    {
+      "epoch": 0.13461214874642435,
+      "grad_norm": 1.5557447509529954,
+      "learning_rate": 2.8843973430085204e-05,
+      "loss": 0.6991,
+      "step": 800
+    },
+    {
+      "epoch": 0.13629480060575466,
+      "grad_norm": 1.9295826624758823,
+      "learning_rate": 2.8812939868470016e-05,
+      "loss": 0.6956,
+      "step": 810
+    },
+    {
+      "epoch": 0.13797745246508497,
+      "grad_norm": 3.3211216557707126,
+      "learning_rate": 2.878151240580548e-05,
+      "loss": 0.6774,
+      "step": 820
+    },
+    {
+      "epoch": 0.13966010432441528,
+      "grad_norm": 4.196064403930616,
+      "learning_rate": 2.874969193830274e-05,
+      "loss": 0.6752,
+      "step": 830
+    },
+    {
+      "epoch": 0.1413427561837456,
+      "grad_norm": 5.574976270137628,
+      "learning_rate": 2.871747937338016e-05,
+      "loss": 0.6553,
+      "step": 840
+    },
+    {
+      "epoch": 0.1430254080430759,
+      "grad_norm": 1.6494038718740478,
+      "learning_rate": 2.8684875629637505e-05,
+      "loss": 0.7152,
+      "step": 850
+    },
+    {
+      "epoch": 0.1447080599024062,
+      "grad_norm": 1.3061892609414858,
+      "learning_rate": 2.8651881636829698e-05,
+      "loss": 0.7462,
+      "step": 860
+    },
+    {
+      "epoch": 0.1463907117617365,
+      "grad_norm": 4.321044418392694,
+      "learning_rate": 2.861849833584032e-05,
+      "loss": 0.6902,
+      "step": 870
+    },
+    {
+      "epoch": 0.1480733636210668,
+      "grad_norm": 2.9444722968009764,
+      "learning_rate": 2.8584726678654787e-05,
+      "loss": 0.6813,
+      "step": 880
+    },
+    {
+      "epoch": 0.1497560154803971,
+      "grad_norm": 1.4940245340163587,
+      "learning_rate": 2.85505676283332e-05,
+      "loss": 0.689,
+      "step": 890
+    },
+    {
+      "epoch": 0.1514386673397274,
+      "grad_norm": 3.3704010040589565,
+      "learning_rate": 2.851602215898287e-05,
+      "loss": 0.6953,
+      "step": 900
+    },
+    {
+      "epoch": 0.15312131919905772,
+      "grad_norm": 1.6597144402924948,
+      "learning_rate": 2.8481091255730552e-05,
+      "loss": 0.7277,
+      "step": 910
+    },
+    {
+      "epoch": 0.15480397105838803,
+      "grad_norm": 10.969872224353953,
+      "learning_rate": 2.844577591469435e-05,
+      "loss": 0.7142,
+      "step": 920
+    },
+    {
+      "epoch": 0.15648662291771834,
+      "grad_norm": 8.45616831264245,
+      "learning_rate": 2.8410077142955304e-05,
+      "loss": 0.7197,
+      "step": 930
+    },
+    {
+      "epoch": 0.15816927477704862,
+      "grad_norm": 2.9594258901214427,
+      "learning_rate": 2.8373995958528683e-05,
+      "loss": 0.7351,
+      "step": 940
+    },
+    {
+      "epoch": 0.15985192663637893,
+      "grad_norm": 2.168676312428759,
+      "learning_rate": 2.8337533390334942e-05,
+      "loss": 0.7544,
+      "step": 950
+    },
+    {
+      "epoch": 0.16153457849570924,
+      "grad_norm": 7.898767360662744,
+      "learning_rate": 2.8300690478170388e-05,
+      "loss": 0.7015,
+      "step": 960
+    },
+    {
+      "epoch": 0.16321723035503954,
+      "grad_norm": 16.83650212945308,
+      "learning_rate": 2.826346827267753e-05,
+      "loss": 0.7139,
+      "step": 970
+    },
+    {
+      "epoch": 0.16489988221436985,
+      "grad_norm": 2.3791337429068977,
+      "learning_rate": 2.8225867835315114e-05,
+      "loss": 0.7053,
+      "step": 980
+    },
+    {
+      "epoch": 0.16658253407370016,
+      "grad_norm": 1.9679363325295285,
+      "learning_rate": 2.8187890238327842e-05,
+      "loss": 0.7313,
+      "step": 990
+    },
+    {
+      "epoch": 0.16826518593303044,
+      "grad_norm": 1.4822625638777076,
+      "learning_rate": 2.814953656471583e-05,
+      "loss": 0.7085,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16994783779236075,
+      "grad_norm": 2.647291447509443,
+      "learning_rate": 2.8110807908203682e-05,
+      "loss": 0.6638,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17163048965169106,
+      "grad_norm": 2.969379719654364,
+      "learning_rate": 2.8071705373209328e-05,
+      "loss": 0.6884,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17331314151102137,
+      "grad_norm": 1.1163745403124403,
+      "learning_rate": 2.803223007481252e-05,
+      "loss": 0.6885,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17499579337035168,
+      "grad_norm": 1.2686557979094786,
+      "learning_rate": 2.7992383138723034e-05,
+      "loss": 0.7037,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17667844522968199,
+      "grad_norm": 4.648945448875594,
+      "learning_rate": 2.7952165701248573e-05,
+      "loss": 0.6933,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1783610970890123,
+      "grad_norm": 4.723564874595428,
+      "learning_rate": 2.7911578909262353e-05,
+      "loss": 0.7144,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18004374894834257,
+      "grad_norm": 5.211806926801946,
+      "learning_rate": 2.787062392017041e-05,
+      "loss": 0.7266,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18172640080767288,
+      "grad_norm": 1.3725560316172503,
+      "learning_rate": 2.7829301901878592e-05,
+      "loss": 0.7445,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1834090526670032,
+      "grad_norm": 0.9012241436004484,
+      "learning_rate": 2.7787614032759243e-05,
+      "loss": 0.6986,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1850917045263335,
+      "grad_norm": 2.912544243603394,
+      "learning_rate": 2.7745561501617605e-05,
+      "loss": 0.7173,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1867743563856638,
+      "grad_norm": 1.4248442614931247,
+      "learning_rate": 2.7703145507657923e-05,
+      "loss": 0.7035,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18845700824499412,
+      "grad_norm": 2.186609904533333,
+      "learning_rate": 2.766036726044926e-05,
+      "loss": 0.7371,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19013966010432443,
+      "grad_norm": 2.0524595532166603,
+      "learning_rate": 2.7617227979890957e-05,
+      "loss": 0.6986,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1918223119636547,
+      "grad_norm": 1.8227045280907195,
+      "learning_rate": 2.7573728896177897e-05,
+      "loss": 0.7075,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19350496382298502,
+      "grad_norm": 1.8425998009576734,
+      "learning_rate": 2.7529871249765397e-05,
+      "loss": 0.6897,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19518761568231532,
+      "grad_norm": 5.3035191638420836,
+      "learning_rate": 2.7485656291333845e-05,
+      "loss": 0.7027,
+      "step": 1160
+    },
+    {
+      "epoch": 0.19687026754164563,
+      "grad_norm": 3.3228474353685504,
+      "learning_rate": 2.7441085281753028e-05,
+      "loss": 0.7091,
+      "step": 1170
+    },
+    {
+      "epoch": 0.19855291940097594,
+      "grad_norm": 3.5016968564731283,
+      "learning_rate": 2.739615949204617e-05,
+      "loss": 0.7241,
+      "step": 1180
+    },
+    {
+      "epoch": 0.20023557126030625,
+      "grad_norm": 1.7190048028902127,
+      "learning_rate": 2.7350880203353703e-05,
+      "loss": 0.7192,
+      "step": 1190
+    },
+    {
+      "epoch": 0.20191822311963656,
+      "grad_norm": 3.7186824247487515,
+      "learning_rate": 2.7305248706896722e-05,
+      "loss": 0.7063,
+      "step": 1200
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 5943,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 400,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.467141180489728e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1200/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e95a8f5e7f8a0f6f3e1f415e9606de2bf6f80315b55f9012ea921093e8d88264
+size 6520

checkpoint-1200/zero_to_fp32.py

@@ -0,0 +1,587 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example: python zero_to_fp32.py . pytorch_model.bin
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+
+    total_files = len(files)
+    state_dicts = []
+    for f in files:
+        state_dict = torch.load(f, map_location=device)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
+        ]
+
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel()
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in param_shapes.items():
+
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+    print(f"Saving fp32 state dict to {output_file}")
+    torch.save(state_dict, output_file)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument(
+        "output_file",
+        type=str,
+        help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, args.output_file, tag=args.tag)

checkpoint-1600/README.md

@@ -0,0 +1,202 @@
+---
+library_name: peft
+base_model: Qwen/Qwen-VL-Chat
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.10.0

checkpoint-1600/adapter_config.json

@@ -0,0 +1,380 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen-VL-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "transformer.h.17.attn.c_proj",
+    "transformer.h.20.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.1.attn.in_proj",
+    "transformer.h.3.attn.c_attn",
+    "transformer.visual.transformer.resblocks.12.attn.in_proj",
+    "transformer.visual.transformer.resblocks.47.attn.in_proj",
+    "transformer.h.28.mlp.w2",
+    "transformer.h.6.mlp.w2",
+    "transformer.h.13.mlp.w1",
+    "transformer.visual.transformer.resblocks.39.attn.out_proj",
+    "transformer.h.2.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.attn.out_proj",
+    "transformer.visual.transformer.resblocks.0.attn.out_proj",
+    "transformer.h.4.attn.c_proj",
+    "transformer.h.22.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.12.attn.out_proj",
+    "transformer.visual.transformer.resblocks.10.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.43.attn.in_proj",
+    "transformer.visual.transformer.resblocks.0.attn.in_proj",
+    "transformer.visual.transformer.resblocks.26.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.11.mlp.c_proj",
+    "transformer.h.0.attn.c_attn",
+    "transformer.h.19.mlp.w2",
+    "transformer.visual.transformer.resblocks.37.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.40.mlp.c_proj",
+    "transformer.h.31.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.mlp.c_fc",
+    "transformer.h.18.mlp.w1",
+    "transformer.h.23.mlp.w2",
+    "transformer.visual.transformer.resblocks.6.attn.out_proj",
+    "transformer.visual.transformer.resblocks.17.attn.in_proj",
+    "transformer.visual.transformer.resblocks.27.attn.out_proj",
+    "transformer.h.12.mlp.w2",
+    "transformer.h.23.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.29.attn.in_proj",
+    "transformer.h.10.mlp.w1",
+    "transformer.visual.transformer.resblocks.18.attn.out_proj",
+    "transformer.visual.transformer.resblocks.4.attn.out_proj",
+    "transformer.visual.transformer.resblocks.8.mlp.c_fc",
+    "transformer.h.9.mlp.w1",
+    "transformer.visual.transformer.resblocks.38.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.attn.in_proj",
+    "transformer.visual.transformer.resblocks.14.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.22.attn.in_proj",
+    "transformer.visual.transformer.resblocks.25.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.23.attn.out_proj",
+    "transformer.visual.transformer.resblocks.3.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.41.mlp.c_proj",
+    "transformer.h.24.attn.c_proj",
+    "transformer.visual.transformer.resblocks.7.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.38.mlp.c_fc",
+    "transformer.h.10.attn.c_attn",
+    "transformer.h.26.attn.c_attn",
+    "transformer.visual.transformer.resblocks.5.attn.in_proj",
+    "transformer.visual.transformer.resblocks.2.attn.out_proj",
+    "transformer.h.7.attn.c_proj",
+    "transformer.h.24.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.34.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.2.mlp.c_proj",
+    "transformer.h.12.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.14.attn.out_proj",
+    "transformer.h.18.attn.c_attn",
+    "transformer.h.23.attn.c_proj",
+    "transformer.h.27.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.26.mlp.c_proj",
+    "transformer.h.3.mlp.w1",
+    "transformer.h.2.mlp.w2",
+    "transformer.visual.transformer.resblocks.45.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.25.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.45.attn.out_proj",
+    "transformer.h.25.mlp.w1",
+    "transformer.visual.transformer.resblocks.15.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.24.attn.in_proj",
+    "transformer.h.1.attn.c_proj",
+    "transformer.h.1.attn.c_attn",
+    "transformer.visual.transformer.resblocks.4.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.9.mlp.c_fc",
+    "transformer.h.13.attn.c_attn",
+    "transformer.visual.transformer.resblocks.40.attn.out_proj",
+    "transformer.h.7.mlp.w2",
+    "transformer.h.9.attn.c_proj",
+    "transformer.h.15.attn.c_attn",
+    "transformer.visual.transformer.resblocks.0.mlp.c_fc",
+    "transformer.h.27.attn.c_attn",
+    "transformer.h.15.mlp.c_proj",
+    "transformer.h.21.mlp.w2",
+    "transformer.h.28.attn.c_proj",
+    "transformer.visual.transformer.resblocks.42.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.16.attn.out_proj",
+    "transformer.h.9.mlp.w2",
+    "transformer.visual.transformer.resblocks.9.attn.in_proj",
+    "transformer.visual.transformer.resblocks.28.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.mlp.c_proj",
+    "transformer.h.11.mlp.w1",
+    "transformer.visual.transformer.resblocks.18.attn.in_proj",
+    "transformer.h.10.attn.c_proj",
+    "transformer.visual.transformer.resblocks.42.mlp.c_fc",
+    "transformer.h.31.attn.c_attn",
+    "transformer.visual.transformer.resblocks.23.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.21.attn.in_proj",
+    "transformer.h.24.mlp.w1",
+    "transformer.visual.transformer.resblocks.35.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.7.mlp.c_proj",
+    "transformer.h.8.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.12.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.7.attn.out_proj",
+    "transformer.h.22.mlp.w2",
+    "transformer.h.29.mlp.w2",
+    "transformer.h.0.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.38.attn.in_proj",
+    "transformer.h.8.mlp.w1",
+    "transformer.h.0.mlp.w1",
+    "transformer.h.26.mlp.w2",
+    "transformer.h.25.attn.c_proj",
+    "transformer.h.27.mlp.w1",
+    "transformer.visual.transformer.resblocks.21.attn.out_proj",
+    "transformer.visual.transformer.resblocks.44.attn.in_proj",
+    "transformer.visual.transformer.resblocks.43.attn.out_proj",
+    "transformer.h.29.attn.c_attn",
+    "transformer.h.24.attn.c_attn",
+    "transformer.visual.transformer.resblocks.17.attn.out_proj",
+    "transformer.h.2.attn.c_proj",
+    "transformer.visual.transformer.resblocks.15.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.11.attn.in_proj",
+    "transformer.visual.transformer.resblocks.17.mlp.c_proj",
+    "transformer.h.11.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.41.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.5.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.18.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.24.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.attn.out_proj",
+    "transformer.h.1.mlp.w2",
+    "transformer.h.21.mlp.c_proj",
+    "transformer.h.23.attn.c_attn",
+    "transformer.visual.transformer.resblocks.34.attn.out_proj",
+    "transformer.h.14.attn.c_attn",
+    "transformer.visual.transformer.resblocks.2.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.31.attn.out_proj",
+    "transformer.visual.transformer.resblocks.30.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.11.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.31.attn.in_proj",
+    "transformer.visual.transformer.resblocks.39.mlp.c_proj",
+    "transformer.h.9.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.20.attn.out_proj",
+    "transformer.h.18.mlp.c_proj",
+    "transformer.h.19.mlp.w1",
+    "transformer.h.9.attn.c_attn",
+    "transformer.visual.transformer.resblocks.36.attn.out_proj",
+    "transformer.visual.transformer.resblocks.7.attn.in_proj",
+    "transformer.visual.transformer.resblocks.30.attn.in_proj",
+    "transformer.visual.transformer.resblocks.47.attn.out_proj",
+    "transformer.visual.transformer.resblocks.0.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.15.attn.in_proj",
+    "transformer.visual.transformer.resblocks.29.attn.out_proj",
+    "transformer.visual.transformer.resblocks.41.attn.in_proj",
+    "transformer.visual.transformer.resblocks.4.attn.in_proj",
+    "transformer.h.25.attn.c_attn",
+    "transformer.visual.transformer.resblocks.12.mlp.c_proj",
+    "transformer.h.16.mlp.w1",
+    "transformer.h.28.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.27.attn.in_proj",
+    "transformer.visual.transformer.resblocks.13.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.attn.in_proj",
+    "transformer.visual.transformer.resblocks.45.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.46.mlp.c_proj",
+    "transformer.h.30.mlp.w1",
+    "transformer.visual.transformer.resblocks.43.mlp.c_fc",
+    "transformer.h.15.mlp.w1",
+    "transformer.h.16.attn.c_proj",
+    "transformer.h.20.mlp.w1",
+    "transformer.visual.transformer.resblocks.21.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.10.mlp.c_proj",
+    "transformer.h.10.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.35.attn.in_proj",
+    "transformer.h.13.mlp.w2",
+    "transformer.visual.transformer.resblocks.8.attn.out_proj",
+    "transformer.visual.transformer.resblocks.20.mlp.c_proj",
+    "transformer.h.22.attn.c_proj",
+    "transformer.h.6.mlp.w1",
+    "transformer.h.18.mlp.w2",
+    "transformer.h.4.mlp.c_proj",
+    "transformer.h.3.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.42.attn.out_proj",
+    "transformer.visual.transformer.resblocks.36.attn.in_proj",
+    "transformer.visual.transformer.resblocks.17.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.43.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.37.attn.in_proj",
+    "transformer.visual.transformer.resblocks.1.attn.out_proj",
+    "transformer.visual.transformer.resblocks.22.mlp.c_fc",
+    "transformer.h.22.mlp.w1",
+    "transformer.visual.transformer.resblocks.44.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.37.attn.out_proj",
+    "transformer.visual.transformer.resblocks.34.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.29.mlp.c_fc",
+    "transformer.h.18.attn.c_proj",
+    "transformer.visual.transformer.resblocks.38.attn.out_proj",
+    "transformer.h.5.attn.c_attn",
+    "transformer.visual.transformer.resblocks.19.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.15.attn.out_proj",
+    "transformer.visual.transformer.resblocks.37.mlp.c_fc",
+    "transformer.h.5.attn.c_proj",
+    "transformer.h.7.attn.c_attn",
+    "transformer.visual.transformer.resblocks.28.attn.out_proj",
+    "transformer.visual.transformer.resblocks.31.mlp.c_proj",
+    "transformer.h.29.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.45.attn.in_proj",
+    "transformer.visual.transformer.resblocks.27.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.10.attn.out_proj",
+    "transformer.visual.transformer.resblocks.40.attn.in_proj",
+    "transformer.h.23.mlp.w1",
+    "transformer.visual.transformer.resblocks.28.attn.in_proj",
+    "transformer.h.12.attn.c_proj",
+    "transformer.h.16.mlp.w2",
+    "transformer.h.27.mlp.w2",
+    "transformer.visual.transformer.resblocks.22.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.9.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.47.mlp.c_proj",
+    "transformer.h.26.attn.c_proj",
+    "transformer.visual.transformer.resblocks.40.mlp.c_fc",
+    "transformer.h.8.mlp.w2",
+    "transformer.visual.transformer.resblocks.27.mlp.c_fc",
+    "transformer.h.17.mlp.w1",
+    "transformer.h.31.mlp.w2",
+    "transformer.visual.transformer.resblocks.11.attn.out_proj",
+    "transformer.h.28.mlp.w1",
+    "transformer.visual.transformer.resblocks.10.attn.in_proj",
+    "transformer.h.12.mlp.w1",
+    "transformer.h.30.mlp.w2",
+    "transformer.visual.transformer.resblocks.13.attn.in_proj",
+    "transformer.h.6.attn.c_attn",
+    "transformer.h.5.mlp.c_proj",
+    "transformer.h.6.mlp.c_proj",
+    "transformer.h.22.attn.c_attn",
+    "transformer.h.13.attn.c_proj",
+    "transformer.visual.transformer.resblocks.46.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.41.attn.out_proj",
+    "transformer.visual.transformer.resblocks.30.mlp.c_fc",
+    "transformer.h.17.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.5.attn.out_proj",
+    "transformer.h.4.mlp.w2",
+    "transformer.visual.transformer.resblocks.1.mlp.c_proj",
+    "transformer.h.11.mlp.w2",
+    "transformer.h.19.attn.c_attn",
+    "transformer.h.14.mlp.w1",
+    "transformer.visual.transformer.resblocks.44.attn.out_proj",
+    "transformer.visual.transformer.resblocks.14.mlp.c_fc",
+    "transformer.h.21.attn.c_attn",
+    "transformer.visual.transformer.resblocks.36.mlp.c_proj",
+    "transformer.h.2.mlp.w1",
+    "transformer.h.14.attn.c_proj",
+    "transformer.visual.transformer.resblocks.46.attn.in_proj",
+    "transformer.h.6.attn.c_proj",
+    "transformer.h.0.mlp.w2",
+    "transformer.h.5.mlp.w1",
+    "transformer.h.30.attn.c_proj",
+    "transformer.h.24.mlp.w2",
+    "transformer.h.0.attn.c_proj",
+    "transformer.visual.transformer.resblocks.4.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.22.attn.out_proj",
+    "transformer.h.10.mlp.w2",
+    "transformer.h.17.mlp.w2",
+    "transformer.visual.transformer.resblocks.23.attn.in_proj",
+    "transformer.visual.transformer.resblocks.36.mlp.c_fc",
+    "transformer.h.20.mlp.w2",
+    "transformer.visual.transformer.resblocks.9.attn.out_proj",
+    "transformer.h.29.mlp.w1",
+    "transformer.visual.transformer.resblocks.20.attn.in_proj",
+    "transformer.visual.transformer.resblocks.20.mlp.c_fc",
+    "transformer.h.15.attn.c_proj",
+    "transformer.h.3.mlp.w2",
+    "transformer.h.30.attn.c_attn",
+    "transformer.visual.transformer.resblocks.47.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.16.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.39.mlp.c_fc",
+    "transformer.h.20.attn.c_attn",
+    "transformer.h.19.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.46.attn.out_proj",
+    "transformer.visual.transformer.resblocks.29.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.19.attn.out_proj",
+    "transformer.visual.transformer.resblocks.26.attn.in_proj",
+    "transformer.visual.transformer.resblocks.16.mlp.c_fc",
+    "transformer.h.11.attn.c_proj",
+    "transformer.h.12.attn.c_attn",
+    "transformer.visual.conv1",
+    "transformer.visual.transformer.resblocks.35.attn.out_proj",
+    "transformer.h.25.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.14.attn.in_proj",
+    "transformer.h.26.mlp.w1",
+    "transformer.visual.transformer.resblocks.1.mlp.c_fc",
+    "transformer.h.7.mlp.c_proj",
+    "transformer.h.29.attn.c_proj",
+    "transformer.h.1.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.mlp.c_proj",
+    "transformer.h.14.mlp.c_proj",
+    "transformer.h.3.attn.c_proj",
+    "transformer.h.25.mlp.w2",
+    "transformer.h.20.attn.c_proj",
+    "transformer.h.16.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.attn.in_proj",
+    "transformer.h.17.attn.c_attn",
+    "transformer.h.14.mlp.w2",
+    "transformer.visual.transformer.resblocks.2.attn.in_proj",
+    "transformer.visual.transformer.resblocks.5.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.33.attn.out_proj",
+    "transformer.h.15.mlp.w2",
+    "transformer.h.4.attn.c_attn",
+    "transformer.h.31.mlp.w1",
+    "transformer.h.11.attn.c_attn",
+    "transformer.visual.transformer.resblocks.23.mlp.c_proj",
+    "transformer.h.7.mlp.w1",
+    "transformer.visual.transformer.resblocks.34.attn.in_proj",
+    "transformer.h.1.mlp.w1",
+    "transformer.visual.transformer.resblocks.28.mlp.c_fc",
+    "transformer.h.21.attn.c_proj",
+    "transformer.h.30.mlp.c_proj",
+    "transformer.h.21.mlp.w1",
+    "transformer.visual.transformer.resblocks.30.attn.out_proj",
+    "transformer.visual.transformer.resblocks.42.attn.in_proj",
+    "transformer.visual.transformer.resblocks.25.attn.out_proj",
+    "transformer.visual.transformer.resblocks.19.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.39.attn.in_proj",
+    "transformer.visual.transformer.resblocks.19.attn.in_proj",
+    "transformer.visual.transformer.resblocks.13.mlp.c_fc",
+    "transformer.h.13.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.25.attn.in_proj",
+    "transformer.visual.transformer.resblocks.31.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.24.attn.out_proj",
+    "transformer.visual.transformer.resblocks.24.mlp.c_fc",
+    "transformer.h.4.mlp.w1",
+    "transformer.h.8.attn.c_attn",
+    "transformer.visual.transformer.resblocks.21.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.44.mlp.c_proj",
+    "transformer.h.28.attn.c_attn",
+    "transformer.visual.transformer.resblocks.18.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.attn.in_proj",
+    "transformer.h.19.attn.c_proj",
+    "transformer.h.2.attn.c_attn",
+    "transformer.visual.transformer.resblocks.35.mlp.c_proj",
+    "transformer.h.26.mlp.c_proj",
+    "transformer.h.8.attn.c_proj",
+    "transformer.h.27.attn.c_proj",
+    "transformer.visual.transformer.resblocks.13.attn.out_proj",
+    "transformer.h.16.attn.c_attn",
+    "transformer.visual.transformer.resblocks.16.attn.in_proj",
+    "transformer.visual.transformer.resblocks.8.attn.in_proj",
+    "transformer.visual.transformer.resblocks.26.attn.out_proj",
+    "transformer.h.31.attn.c_proj",
+    "transformer.h.5.mlp.w2",
+    "transformer.visual.transformer.resblocks.8.mlp.c_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-1600/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4132637e7818ab04e79770c537ecce329ad66d75a5084c122da7b6e24018491d
+size 469105640

checkpoint-1600/latest

@@ -0,0 +1 @@
+global_step1600

checkpoint-1600/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0943be1d73f8cd864a8d86cf602e01a0ee9483c4bbd1287ca1aa6a70a07a7d78
+size 14960

checkpoint-1600/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1814e986cf45968069746e25fb44856688645b26635b32a478bed3330978b28
+size 14960

checkpoint-1600/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bfdaf471dc63d05f3a4e9aff471d87f770511c8529541d9f1d14a82ae9e16fd9
+size 14960

checkpoint-1600/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8557ae681bc439c82fd4086d0ae897ebef2a086d2acbb923505fe7e63067cbd2
+size 14960

checkpoint-1600/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0d0794962fd5d327d9bc0d61b8681692b6df3d1030d5844836ce2192acb24b7
+size 1064

checkpoint-1600/special_tokens_map.json

@@ -0,0 +1,3 @@
+{
+  "pad_token": "<|endoftext|>"
+}

checkpoint-1600/tokenizer_config.json

@@ -0,0 +1,14 @@
+{
+  "added_tokens_decoder": {},
+  "auto_map": {
+    "AutoTokenizer": [
+      "Qwen/Qwen-VL-Chat--tokenization_qwen.QWenTokenizer",
+      null
+    ]
+  },
+  "clean_up_tokenization_spaces": true,
+  "model_max_length": 1280,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "tokenizer_class": "QWenTokenizer"
+}

checkpoint-1600/trainer_state.json

@@ -0,0 +1,1153 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.2692242974928487,
+  "eval_steps": 500,
+  "global_step": 1600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016826518593303045,
+      "grad_norm": 5.367858933563703,
+      "learning_rate": 4.9999999999999996e-06,
+      "loss": 0.9537,
+      "step": 10
+    },
+    {
+      "epoch": 0.003365303718660609,
+      "grad_norm": 9.386746384686745,
+      "learning_rate": 9.999999999999999e-06,
+      "loss": 0.943,
+      "step": 20
+    },
+    {
+      "epoch": 0.005047955577990914,
+      "grad_norm": 7.387362447577942,
+      "learning_rate": 1.5e-05,
+      "loss": 0.934,
+      "step": 30
+    },
+    {
+      "epoch": 0.006730607437321218,
+      "grad_norm": 6.9256319824932655,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 0.8376,
+      "step": 40
+    },
+    {
+      "epoch": 0.008413259296651522,
+      "grad_norm": 9.1148382590838,
+      "learning_rate": 2.5e-05,
+      "loss": 0.8484,
+      "step": 50
+    },
+    {
+      "epoch": 0.010095911155981827,
+      "grad_norm": 3.9989232759892426,
+      "learning_rate": 3e-05,
+      "loss": 0.8097,
+      "step": 60
+    },
+    {
+      "epoch": 0.011778563015312132,
+      "grad_norm": 3.892371218590039,
+      "learning_rate": 2.9999786123888308e-05,
+      "loss": 0.7811,
+      "step": 70
+    },
+    {
+      "epoch": 0.013461214874642436,
+      "grad_norm": 8.096662196282066,
+      "learning_rate": 2.9999144501652298e-05,
+      "loss": 0.7446,
+      "step": 80
+    },
+    {
+      "epoch": 0.01514386673397274,
+      "grad_norm": 1.5769306611206149,
+      "learning_rate": 2.9998075151588992e-05,
+      "loss": 0.7258,
+      "step": 90
+    },
+    {
+      "epoch": 0.016826518593303044,
+      "grad_norm": 8.47430485487969,
+      "learning_rate": 2.999657810419285e-05,
+      "loss": 0.7052,
+      "step": 100
+    },
+    {
+      "epoch": 0.01850917045263335,
+      "grad_norm": 2.363071299913598,
+      "learning_rate": 2.999465340215489e-05,
+      "loss": 0.7657,
+      "step": 110
+    },
+    {
+      "epoch": 0.020191822311963654,
+      "grad_norm": 1.9252385425154874,
+      "learning_rate": 2.999230110036149e-05,
+      "loss": 0.7329,
+      "step": 120
+    },
+    {
+      "epoch": 0.02187447417129396,
+      "grad_norm": 8.946028475031488,
+      "learning_rate": 2.99895212658928e-05,
+      "loss": 0.7304,
+      "step": 130
+    },
+    {
+      "epoch": 0.023557126030624265,
+      "grad_norm": 6.877609312630206,
+      "learning_rate": 2.9986313978020846e-05,
+      "loss": 0.7453,
+      "step": 140
+    },
+    {
+      "epoch": 0.02523977788995457,
+      "grad_norm": 2.5256324882367993,
+      "learning_rate": 2.9982679328207262e-05,
+      "loss": 0.7366,
+      "step": 150
+    },
+    {
+      "epoch": 0.02692242974928487,
+      "grad_norm": 2.709550398238738,
+      "learning_rate": 2.9978617420100692e-05,
+      "loss": 0.7258,
+      "step": 160
+    },
+    {
+      "epoch": 0.028605081608615177,
+      "grad_norm": 1.543550019689774,
+      "learning_rate": 2.9974128369533805e-05,
+      "loss": 0.7372,
+      "step": 170
+    },
+    {
+      "epoch": 0.03028773346794548,
+      "grad_norm": 3.3453966881155504,
+      "learning_rate": 2.9969212304520034e-05,
+      "loss": 0.743,
+      "step": 180
+    },
+    {
+      "epoch": 0.03197038532727579,
+      "grad_norm": 1.922001656181265,
+      "learning_rate": 2.9963869365249895e-05,
+      "loss": 0.7819,
+      "step": 190
+    },
+    {
+      "epoch": 0.03365303718660609,
+      "grad_norm": 2.0611188483400036,
+      "learning_rate": 2.995809970408699e-05,
+      "loss": 0.7155,
+      "step": 200
+    },
+    {
+      "epoch": 0.0353356890459364,
+      "grad_norm": 1.5313041833127259,
+      "learning_rate": 2.9951903485563685e-05,
+      "loss": 0.7322,
+      "step": 210
+    },
+    {
+      "epoch": 0.0370183409052667,
+      "grad_norm": 2.0124191694435085,
+      "learning_rate": 2.99452808863764e-05,
+      "loss": 0.6759,
+      "step": 220
+    },
+    {
+      "epoch": 0.03870099276459701,
+      "grad_norm": 3.182123324389477,
+      "learning_rate": 2.993823209538056e-05,
+      "loss": 0.6953,
+      "step": 230
+    },
+    {
+      "epoch": 0.04038364462392731,
+      "grad_norm": 1.6122782177661379,
+      "learning_rate": 2.9930757313585238e-05,
+      "loss": 0.6953,
+      "step": 240
+    },
+    {
+      "epoch": 0.04206629648325761,
+      "grad_norm": 2.2027482596695647,
+      "learning_rate": 2.9922856754147406e-05,
+      "loss": 0.7301,
+      "step": 250
+    },
+    {
+      "epoch": 0.04374894834258792,
+      "grad_norm": 2.6782477155989213,
+      "learning_rate": 2.9914530642365852e-05,
+      "loss": 0.6891,
+      "step": 260
+    },
+    {
+      "epoch": 0.04543160020191822,
+      "grad_norm": 1.9740401144541417,
+      "learning_rate": 2.990577921567476e-05,
+      "loss": 0.7231,
+      "step": 270
+    },
+    {
+      "epoch": 0.04711425206124853,
+      "grad_norm": 1.719874620968932,
+      "learning_rate": 2.989660272363696e-05,
+      "loss": 0.7505,
+      "step": 280
+    },
+    {
+      "epoch": 0.04879690392057883,
+      "grad_norm": 1.3138364164203409,
+      "learning_rate": 2.988700142793676e-05,
+      "loss": 0.7116,
+      "step": 290
+    },
+    {
+      "epoch": 0.05047955577990914,
+      "grad_norm": 5.853627389344256,
+      "learning_rate": 2.9876975602372536e-05,
+      "loss": 0.719,
+      "step": 300
+    },
+    {
+      "epoch": 0.05216220763923944,
+      "grad_norm": 2.347259437170711,
+      "learning_rate": 2.9866525532848906e-05,
+      "loss": 0.6803,
+      "step": 310
+    },
+    {
+      "epoch": 0.05384485949856974,
+      "grad_norm": 1.937679220955038,
+      "learning_rate": 2.9855651517368567e-05,
+      "loss": 0.7461,
+      "step": 320
+    },
+    {
+      "epoch": 0.05552751135790005,
+      "grad_norm": 1.6661300351569575,
+      "learning_rate": 2.9844353866023802e-05,
+      "loss": 0.7472,
+      "step": 330
+    },
+    {
+      "epoch": 0.05721016321723035,
+      "grad_norm": 2.357915869204484,
+      "learning_rate": 2.9832632900987642e-05,
+      "loss": 0.7148,
+      "step": 340
+    },
+    {
+      "epoch": 0.05889281507656066,
+      "grad_norm": 4.398815186243292,
+      "learning_rate": 2.982048895650468e-05,
+      "loss": 0.6992,
+      "step": 350
+    },
+    {
+      "epoch": 0.06057546693589096,
+      "grad_norm": 12.662682224480092,
+      "learning_rate": 2.9807922378881537e-05,
+      "loss": 0.7539,
+      "step": 360
+    },
+    {
+      "epoch": 0.06225811879522127,
+      "grad_norm": 0.8642696401357872,
+      "learning_rate": 2.979493352647697e-05,
+      "loss": 0.7212,
+      "step": 370
+    },
+    {
+      "epoch": 0.06394077065455157,
+      "grad_norm": 27.047937858232604,
+      "learning_rate": 2.9781522769691686e-05,
+      "loss": 0.722,
+      "step": 380
+    },
+    {
+      "epoch": 0.06562342251388188,
+      "grad_norm": 2.598805292448644,
+      "learning_rate": 2.9767690490957758e-05,
+      "loss": 0.7065,
+      "step": 390
+    },
+    {
+      "epoch": 0.06730607437321218,
+      "grad_norm": 1.2314762895092763,
+      "learning_rate": 2.9753437084727713e-05,
+      "loss": 0.7498,
+      "step": 400
+    },
+    {
+      "epoch": 0.06898872623254249,
+      "grad_norm": 1.6421909669790502,
+      "learning_rate": 2.9738762957463292e-05,
+      "loss": 0.6992,
+      "step": 410
+    },
+    {
+      "epoch": 0.0706713780918728,
+      "grad_norm": 2.023552968622588,
+      "learning_rate": 2.9723668527623877e-05,
+      "loss": 0.6943,
+      "step": 420
+    },
+    {
+      "epoch": 0.0723540299512031,
+      "grad_norm": 1.5172337910969138,
+      "learning_rate": 2.9708154225654526e-05,
+      "loss": 0.6987,
+      "step": 430
+    },
+    {
+      "epoch": 0.0740366818105334,
+      "grad_norm": 1.197852135730745,
+      "learning_rate": 2.9692220493973712e-05,
+      "loss": 0.7302,
+      "step": 440
+    },
+    {
+      "epoch": 0.0757193336698637,
+      "grad_norm": 2.4396443837967183,
+      "learning_rate": 2.9675867786960718e-05,
+      "loss": 0.7318,
+      "step": 450
+    },
+    {
+      "epoch": 0.07740198552919401,
+      "grad_norm": 1.4599851880563282,
+      "learning_rate": 2.9659096570942654e-05,
+      "loss": 0.6941,
+      "step": 460
+    },
+    {
+      "epoch": 0.07908463738852431,
+      "grad_norm": 1.117755825364562,
+      "learning_rate": 2.9641907324181194e-05,
+      "loss": 0.7399,
+      "step": 470
+    },
+    {
+      "epoch": 0.08076728924785462,
+      "grad_norm": 2.9235378164576242,
+      "learning_rate": 2.96243005368589e-05,
+      "loss": 0.7207,
+      "step": 480
+    },
+    {
+      "epoch": 0.08244994110718493,
+      "grad_norm": 7.308883163781362,
+      "learning_rate": 2.960627671106527e-05,
+      "loss": 0.682,
+      "step": 490
+    },
+    {
+      "epoch": 0.08413259296651522,
+      "grad_norm": 3.4394827932955234,
+      "learning_rate": 2.9587836360782405e-05,
+      "loss": 0.708,
+      "step": 500
+    },
+    {
+      "epoch": 0.08581524482584553,
+      "grad_norm": 3.2314529856927634,
+      "learning_rate": 2.9568980011870357e-05,
+      "loss": 0.7335,
+      "step": 510
+    },
+    {
+      "epoch": 0.08749789668517584,
+      "grad_norm": 1.825724533695325,
+      "learning_rate": 2.954970820205214e-05,
+      "loss": 0.6951,
+      "step": 520
+    },
+    {
+      "epoch": 0.08918054854450615,
+      "grad_norm": 3.3231741746640076,
+      "learning_rate": 2.9530021480898393e-05,
+      "loss": 0.7793,
+      "step": 530
+    },
+    {
+      "epoch": 0.09086320040383644,
+      "grad_norm": 1.3097651462571123,
+      "learning_rate": 2.9509920409811696e-05,
+      "loss": 0.7087,
+      "step": 540
+    },
+    {
+      "epoch": 0.09254585226316675,
+      "grad_norm": 6.685911471215255,
+      "learning_rate": 2.9489405562010565e-05,
+      "loss": 0.6906,
+      "step": 550
+    },
+    {
+      "epoch": 0.09422850412249706,
+      "grad_norm": 2.870746617513948,
+      "learning_rate": 2.9468477522513132e-05,
+      "loss": 0.7028,
+      "step": 560
+    },
+    {
+      "epoch": 0.09591115598182735,
+      "grad_norm": 1.782555352805469,
+      "learning_rate": 2.9447136888120408e-05,
+      "loss": 0.6901,
+      "step": 570
+    },
+    {
+      "epoch": 0.09759380784115766,
+      "grad_norm": 2.336519711000487,
+      "learning_rate": 2.9425384267399327e-05,
+      "loss": 0.7779,
+      "step": 580
+    },
+    {
+      "epoch": 0.09927645970048797,
+      "grad_norm": 8.935574410818228,
+      "learning_rate": 2.940322028066534e-05,
+      "loss": 0.7503,
+      "step": 590
+    },
+    {
+      "epoch": 0.10095911155981828,
+      "grad_norm": 2.754713786882031,
+      "learning_rate": 2.938064555996476e-05,
+      "loss": 0.7208,
+      "step": 600
+    },
+    {
+      "epoch": 0.10264176341914857,
+      "grad_norm": 1.5082503557652136,
+      "learning_rate": 2.9357660749056713e-05,
+      "loss": 0.7169,
+      "step": 610
+    },
+    {
+      "epoch": 0.10432441527847888,
+      "grad_norm": 9.04522194526273,
+      "learning_rate": 2.9334266503394803e-05,
+      "loss": 0.6927,
+      "step": 620
+    },
+    {
+      "epoch": 0.10600706713780919,
+      "grad_norm": 55.28278686388287,
+      "learning_rate": 2.9310463490108397e-05,
+      "loss": 0.7107,
+      "step": 630
+    },
+    {
+      "epoch": 0.10768971899713949,
+      "grad_norm": 3.721916069105249,
+      "learning_rate": 2.928625238798362e-05,
+      "loss": 0.6951,
+      "step": 640
+    },
+    {
+      "epoch": 0.1093723708564698,
+      "grad_norm": 2.5040797323750112,
+      "learning_rate": 2.9261633887443993e-05,
+      "loss": 0.6916,
+      "step": 650
+    },
+    {
+      "epoch": 0.1110550227158001,
+      "grad_norm": 3.5468924769840617,
+      "learning_rate": 2.9236608690530738e-05,
+      "loss": 0.7077,
+      "step": 660
+    },
+    {
+      "epoch": 0.11273767457513041,
+      "grad_norm": 3.0266819778200746,
+      "learning_rate": 2.921117751088276e-05,
+      "loss": 0.6952,
+      "step": 670
+    },
+    {
+      "epoch": 0.1144203264344607,
+      "grad_norm": 1.634743894298146,
+      "learning_rate": 2.91853410737163e-05,
+      "loss": 0.6936,
+      "step": 680
+    },
+    {
+      "epoch": 0.11610297829379101,
+      "grad_norm": 1.0925365801520501,
+      "learning_rate": 2.915910011580426e-05,
+      "loss": 0.7317,
+      "step": 690
+    },
+    {
+      "epoch": 0.11778563015312132,
+      "grad_norm": 1.6959112138540386,
+      "learning_rate": 2.9132455385455176e-05,
+      "loss": 0.6917,
+      "step": 700
+    },
+    {
+      "epoch": 0.11946828201245162,
+      "grad_norm": 1.9723433746891168,
+      "learning_rate": 2.9105407642491895e-05,
+      "loss": 0.7209,
+      "step": 710
+    },
+    {
+      "epoch": 0.12115093387178193,
+      "grad_norm": 2.1537215293733833,
+      "learning_rate": 2.907795765822989e-05,
+      "loss": 0.7488,
+      "step": 720
+    },
+    {
+      "epoch": 0.12283358573111224,
+      "grad_norm": 3.227101869737169,
+      "learning_rate": 2.9050106215455283e-05,
+      "loss": 0.7152,
+      "step": 730
+    },
+    {
+      "epoch": 0.12451623759044254,
+      "grad_norm": 2.7222358893572554,
+      "learning_rate": 2.9021854108402516e-05,
+      "loss": 0.708,
+      "step": 740
+    },
+    {
+      "epoch": 0.12619888944977284,
+      "grad_norm": 2.1054843767538136,
+      "learning_rate": 2.8993202142731693e-05,
+      "loss": 0.7251,
+      "step": 750
+    },
+    {
+      "epoch": 0.12788154130910315,
+      "grad_norm": 2.11845883419618,
+      "learning_rate": 2.8964151135505616e-05,
+      "loss": 0.7405,
+      "step": 760
+    },
+    {
+      "epoch": 0.12956419316843346,
+      "grad_norm": 13.171512404187755,
+      "learning_rate": 2.8934701915166477e-05,
+      "loss": 0.6844,
+      "step": 770
+    },
+    {
+      "epoch": 0.13124684502776376,
+      "grad_norm": 2.7633375632879127,
+      "learning_rate": 2.890485532151225e-05,
+      "loss": 0.6766,
+      "step": 780
+    },
+    {
+      "epoch": 0.13292949688709407,
+      "grad_norm": 1.8420785342693768,
+      "learning_rate": 2.887461220567271e-05,
+      "loss": 0.7037,
+      "step": 790
+    },
+    {
+      "epoch": 0.13461214874642435,
+      "grad_norm": 1.5557447509529954,
+      "learning_rate": 2.8843973430085204e-05,
+      "loss": 0.6991,
+      "step": 800
+    },
+    {
+      "epoch": 0.13629480060575466,
+      "grad_norm": 1.9295826624758823,
+      "learning_rate": 2.8812939868470016e-05,
+      "loss": 0.6956,
+      "step": 810
+    },
+    {
+      "epoch": 0.13797745246508497,
+      "grad_norm": 3.3211216557707126,
+      "learning_rate": 2.878151240580548e-05,
+      "loss": 0.6774,
+      "step": 820
+    },
+    {
+      "epoch": 0.13966010432441528,
+      "grad_norm": 4.196064403930616,
+      "learning_rate": 2.874969193830274e-05,
+      "loss": 0.6752,
+      "step": 830
+    },
+    {
+      "epoch": 0.1413427561837456,
+      "grad_norm": 5.574976270137628,
+      "learning_rate": 2.871747937338016e-05,
+      "loss": 0.6553,
+      "step": 840
+    },
+    {
+      "epoch": 0.1430254080430759,
+      "grad_norm": 1.6494038718740478,
+      "learning_rate": 2.8684875629637505e-05,
+      "loss": 0.7152,
+      "step": 850
+    },
+    {
+      "epoch": 0.1447080599024062,
+      "grad_norm": 1.3061892609414858,
+      "learning_rate": 2.8651881636829698e-05,
+      "loss": 0.7462,
+      "step": 860
+    },
+    {
+      "epoch": 0.1463907117617365,
+      "grad_norm": 4.321044418392694,
+      "learning_rate": 2.861849833584032e-05,
+      "loss": 0.6902,
+      "step": 870
+    },
+    {
+      "epoch": 0.1480733636210668,
+      "grad_norm": 2.9444722968009764,
+      "learning_rate": 2.8584726678654787e-05,
+      "loss": 0.6813,
+      "step": 880
+    },
+    {
+      "epoch": 0.1497560154803971,
+      "grad_norm": 1.4940245340163587,
+      "learning_rate": 2.85505676283332e-05,
+      "loss": 0.689,
+      "step": 890
+    },
+    {
+      "epoch": 0.1514386673397274,
+      "grad_norm": 3.3704010040589565,
+      "learning_rate": 2.851602215898287e-05,
+      "loss": 0.6953,
+      "step": 900
+    },
+    {
+      "epoch": 0.15312131919905772,
+      "grad_norm": 1.6597144402924948,
+      "learning_rate": 2.8481091255730552e-05,
+      "loss": 0.7277,
+      "step": 910
+    },
+    {
+      "epoch": 0.15480397105838803,
+      "grad_norm": 10.969872224353953,
+      "learning_rate": 2.844577591469435e-05,
+      "loss": 0.7142,
+      "step": 920
+    },
+    {
+      "epoch": 0.15648662291771834,
+      "grad_norm": 8.45616831264245,
+      "learning_rate": 2.8410077142955304e-05,
+      "loss": 0.7197,
+      "step": 930
+    },
+    {
+      "epoch": 0.15816927477704862,
+      "grad_norm": 2.9594258901214427,
+      "learning_rate": 2.8373995958528683e-05,
+      "loss": 0.7351,
+      "step": 940
+    },
+    {
+      "epoch": 0.15985192663637893,
+      "grad_norm": 2.168676312428759,
+      "learning_rate": 2.8337533390334942e-05,
+      "loss": 0.7544,
+      "step": 950
+    },
+    {
+      "epoch": 0.16153457849570924,
+      "grad_norm": 7.898767360662744,
+      "learning_rate": 2.8300690478170388e-05,
+      "loss": 0.7015,
+      "step": 960
+    },
+    {
+      "epoch": 0.16321723035503954,
+      "grad_norm": 16.83650212945308,
+      "learning_rate": 2.826346827267753e-05,
+      "loss": 0.7139,
+      "step": 970
+    },
+    {
+      "epoch": 0.16489988221436985,
+      "grad_norm": 2.3791337429068977,
+      "learning_rate": 2.8225867835315114e-05,
+      "loss": 0.7053,
+      "step": 980
+    },
+    {
+      "epoch": 0.16658253407370016,
+      "grad_norm": 1.9679363325295285,
+      "learning_rate": 2.8187890238327842e-05,
+      "loss": 0.7313,
+      "step": 990
+    },
+    {
+      "epoch": 0.16826518593303044,
+      "grad_norm": 1.4822625638777076,
+      "learning_rate": 2.814953656471583e-05,
+      "loss": 0.7085,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16994783779236075,
+      "grad_norm": 2.647291447509443,
+      "learning_rate": 2.8110807908203682e-05,
+      "loss": 0.6638,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17163048965169106,
+      "grad_norm": 2.969379719654364,
+      "learning_rate": 2.8071705373209328e-05,
+      "loss": 0.6884,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17331314151102137,
+      "grad_norm": 1.1163745403124403,
+      "learning_rate": 2.803223007481252e-05,
+      "loss": 0.6885,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17499579337035168,
+      "grad_norm": 1.2686557979094786,
+      "learning_rate": 2.7992383138723034e-05,
+      "loss": 0.7037,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17667844522968199,
+      "grad_norm": 4.648945448875594,
+      "learning_rate": 2.7952165701248573e-05,
+      "loss": 0.6933,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1783610970890123,
+      "grad_norm": 4.723564874595428,
+      "learning_rate": 2.7911578909262353e-05,
+      "loss": 0.7144,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18004374894834257,
+      "grad_norm": 5.211806926801946,
+      "learning_rate": 2.787062392017041e-05,
+      "loss": 0.7266,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18172640080767288,
+      "grad_norm": 1.3725560316172503,
+      "learning_rate": 2.7829301901878592e-05,
+      "loss": 0.7445,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1834090526670032,
+      "grad_norm": 0.9012241436004484,
+      "learning_rate": 2.7787614032759243e-05,
+      "loss": 0.6986,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1850917045263335,
+      "grad_norm": 2.912544243603394,
+      "learning_rate": 2.7745561501617605e-05,
+      "loss": 0.7173,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1867743563856638,
+      "grad_norm": 1.4248442614931247,
+      "learning_rate": 2.7703145507657923e-05,
+      "loss": 0.7035,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18845700824499412,
+      "grad_norm": 2.186609904533333,
+      "learning_rate": 2.766036726044926e-05,
+      "loss": 0.7371,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19013966010432443,
+      "grad_norm": 2.0524595532166603,
+      "learning_rate": 2.7617227979890957e-05,
+      "loss": 0.6986,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1918223119636547,
+      "grad_norm": 1.8227045280907195,
+      "learning_rate": 2.7573728896177897e-05,
+      "loss": 0.7075,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19350496382298502,
+      "grad_norm": 1.8425998009576734,
+      "learning_rate": 2.7529871249765397e-05,
+      "loss": 0.6897,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19518761568231532,
+      "grad_norm": 5.3035191638420836,
+      "learning_rate": 2.7485656291333845e-05,
+      "loss": 0.7027,
+      "step": 1160
+    },
+    {
+      "epoch": 0.19687026754164563,
+      "grad_norm": 3.3228474353685504,
+      "learning_rate": 2.7441085281753028e-05,
+      "loss": 0.7091,
+      "step": 1170
+    },
+    {
+      "epoch": 0.19855291940097594,
+      "grad_norm": 3.5016968564731283,
+      "learning_rate": 2.739615949204617e-05,
+      "loss": 0.7241,
+      "step": 1180
+    },
+    {
+      "epoch": 0.20023557126030625,
+      "grad_norm": 1.7190048028902127,
+      "learning_rate": 2.7350880203353703e-05,
+      "loss": 0.7192,
+      "step": 1190
+    },
+    {
+      "epoch": 0.20191822311963656,
+      "grad_norm": 3.7186824247487515,
+      "learning_rate": 2.7305248706896722e-05,
+      "loss": 0.7063,
+      "step": 1200
+    },
+    {
+      "epoch": 0.20360087497896684,
+      "grad_norm": 4.1717869895766935,
+      "learning_rate": 2.7259266303940164e-05,
+      "loss": 0.7088,
+      "step": 1210
+    },
+    {
+      "epoch": 0.20528352683829715,
+      "grad_norm": 2.5124857963805804,
+      "learning_rate": 2.7212934305755697e-05,
+      "loss": 0.7198,
+      "step": 1220
+    },
+    {
+      "epoch": 0.20696617869762746,
+      "grad_norm": 2.095136268936366,
+      "learning_rate": 2.7166254033584343e-05,
+      "loss": 0.753,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20864883055695777,
+      "grad_norm": 3.2661098868577256,
+      "learning_rate": 2.7119226818598784e-05,
+      "loss": 0.6779,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21033148241628807,
+      "grad_norm": 3.055506603735091,
+      "learning_rate": 2.7071854001865402e-05,
+      "loss": 0.7013,
+      "step": 1250
+    },
+    {
+      "epoch": 0.21201413427561838,
+      "grad_norm": 12.522953778477769,
+      "learning_rate": 2.702413693430604e-05,
+      "loss": 0.7088,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2136967861349487,
+      "grad_norm": 3.476240301739368,
+      "learning_rate": 2.697607697665948e-05,
+      "loss": 0.689,
+      "step": 1270
+    },
+    {
+      "epoch": 0.21537943799427897,
+      "grad_norm": 1.1862686197570156,
+      "learning_rate": 2.6927675499442648e-05,
+      "loss": 0.7243,
+      "step": 1280
+    },
+    {
+      "epoch": 0.21706208985360928,
+      "grad_norm": 1.6505042403801382,
+      "learning_rate": 2.68789338829115e-05,
+      "loss": 0.7083,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2187447417129396,
+      "grad_norm": 4.74071740077375,
+      "learning_rate": 2.6829853517021698e-05,
+      "loss": 0.7016,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2204273935722699,
+      "grad_norm": 4.124079283639458,
+      "learning_rate": 2.6780435801388945e-05,
+      "loss": 0.7077,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2221100454316002,
+      "grad_norm": 1.9487864410536297,
+      "learning_rate": 2.6730682145249093e-05,
+      "loss": 0.7355,
+      "step": 1320
+    },
+    {
+      "epoch": 0.22379269729093051,
+      "grad_norm": 2.4839241050514733,
+      "learning_rate": 2.668059396741795e-05,
+      "loss": 0.7092,
+      "step": 1330
+    },
+    {
+      "epoch": 0.22547534915026082,
+      "grad_norm": 2.841913657394254,
+      "learning_rate": 2.6630172696250804e-05,
+      "loss": 0.7303,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2271580010095911,
+      "grad_norm": 2.7442870185873347,
+      "learning_rate": 2.6579419769601715e-05,
+      "loss": 0.6739,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2288406528689214,
+      "grad_norm": 1.3854365909071105,
+      "learning_rate": 2.6528336634782493e-05,
+      "loss": 0.7073,
+      "step": 1360
+    },
+    {
+      "epoch": 0.23052330472825172,
+      "grad_norm": 3.115941001607779,
+      "learning_rate": 2.6476924748521443e-05,
+      "loss": 0.7267,
+      "step": 1370
+    },
+    {
+      "epoch": 0.23220595658758203,
+      "grad_norm": 6.9185951332741,
+      "learning_rate": 2.6425185576921812e-05,
+      "loss": 0.7456,
+      "step": 1380
+    },
+    {
+      "epoch": 0.23388860844691234,
+      "grad_norm": 2.378601355345996,
+      "learning_rate": 2.637312059541997e-05,
+      "loss": 0.6912,
+      "step": 1390
+    },
+    {
+      "epoch": 0.23557126030624265,
+      "grad_norm": 2.7929947858543906,
+      "learning_rate": 2.632073128874336e-05,
+      "loss": 0.7184,
+      "step": 1400
+    },
+    {
+      "epoch": 0.23725391216557296,
+      "grad_norm": 1.5382855773213957,
+      "learning_rate": 2.6268019150868144e-05,
+      "loss": 0.7099,
+      "step": 1410
+    },
+    {
+      "epoch": 0.23893656402490324,
+      "grad_norm": 6.1010563795570025,
+      "learning_rate": 2.62149856849766e-05,
+      "loss": 0.6895,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24061921588423354,
+      "grad_norm": 5.999491987974443,
+      "learning_rate": 2.616163240341426e-05,
+      "loss": 0.7493,
+      "step": 1430
+    },
+    {
+      "epoch": 0.24230186774356385,
+      "grad_norm": 2.837037600849311,
+      "learning_rate": 2.6107960827646774e-05,
+      "loss": 0.7176,
+      "step": 1440
+    },
+    {
+      "epoch": 0.24398451960289416,
+      "grad_norm": 1.7029089834427125,
+      "learning_rate": 2.6053972488216538e-05,
+      "loss": 0.6852,
+      "step": 1450
+    },
+    {
+      "epoch": 0.24566717146222447,
+      "grad_norm": 1.382189249222589,
+      "learning_rate": 2.5999668924699035e-05,
+      "loss": 0.685,
+      "step": 1460
+    },
+    {
+      "epoch": 0.24734982332155478,
+      "grad_norm": 1.9496045543050813,
+      "learning_rate": 2.5945051685658923e-05,
+      "loss": 0.6591,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2490324751808851,
+      "grad_norm": 5.479390805764353,
+      "learning_rate": 2.5890122328605908e-05,
+      "loss": 0.7085,
+      "step": 1480
+    },
+    {
+      "epoch": 0.25071512704021537,
+      "grad_norm": 1.7567995670915637,
+      "learning_rate": 2.5834882419950295e-05,
+      "loss": 0.7091,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2523977788995457,
+      "grad_norm": 1.9685911084195309,
+      "learning_rate": 2.577933353495833e-05,
+      "loss": 0.7218,
+      "step": 1500
+    },
+    {
+      "epoch": 0.254080430758876,
+      "grad_norm": 3.400633915540874,
+      "learning_rate": 2.5723477257707293e-05,
+      "loss": 0.7148,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2557630826182063,
+      "grad_norm": 1.2116738326443663,
+      "learning_rate": 2.566731518104029e-05,
+      "loss": 0.7321,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2574457344775366,
+      "grad_norm": 1.3376343864594256,
+      "learning_rate": 2.5610848906520878e-05,
+      "loss": 0.748,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2591283863368669,
+      "grad_norm": 2.6089861003232055,
+      "learning_rate": 2.5554080044387344e-05,
+      "loss": 0.7127,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2608110381961972,
+      "grad_norm": 3.2047926120640526,
+      "learning_rate": 2.5497010213506825e-05,
+      "loss": 0.7262,
+      "step": 1550
+    },
+    {
+      "epoch": 0.26249369005552753,
+      "grad_norm": 1.4899957348295265,
+      "learning_rate": 2.5439641041329128e-05,
+      "loss": 0.7122,
+      "step": 1560
+    },
+    {
+      "epoch": 0.26417634191485784,
+      "grad_norm": 3.595968473922136,
+      "learning_rate": 2.5381974163840313e-05,
+      "loss": 0.7092,
+      "step": 1570
+    },
+    {
+      "epoch": 0.26585899377418815,
+      "grad_norm": 3.5232117574234003,
+      "learning_rate": 2.532401122551605e-05,
+      "loss": 0.6924,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2675416456335184,
+      "grad_norm": 2.618947453668302,
+      "learning_rate": 2.526575387927473e-05,
+      "loss": 0.7067,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2692242974928487,
+      "grad_norm": 3.6282673284589566,
+      "learning_rate": 2.52072037864303e-05,
+      "loss": 0.6945,
+      "step": 1600
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 5943,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 400,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.289521573986304e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1600/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e95a8f5e7f8a0f6f3e1f415e9606de2bf6f80315b55f9012ea921093e8d88264
+size 6520

checkpoint-1600/zero_to_fp32.py

@@ -0,0 +1,587 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example: python zero_to_fp32.py . pytorch_model.bin
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+
+    total_files = len(files)
+    state_dicts = []
+    for f in files:
+        state_dict = torch.load(f, map_location=device)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
+        ]
+
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel()
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in param_shapes.items():
+
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+    print(f"Saving fp32 state dict to {output_file}")
+    torch.save(state_dict, output_file)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument(
+        "output_file",
+        type=str,
+        help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, args.output_file, tag=args.tag)

checkpoint-2000/README.md

@@ -0,0 +1,202 @@
+---
+library_name: peft
+base_model: Qwen/Qwen-VL-Chat
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.10.0

checkpoint-2000/adapter_config.json

@@ -0,0 +1,380 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen-VL-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "transformer.h.17.attn.c_proj",
+    "transformer.h.20.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.1.attn.in_proj",
+    "transformer.h.3.attn.c_attn",
+    "transformer.visual.transformer.resblocks.12.attn.in_proj",
+    "transformer.visual.transformer.resblocks.47.attn.in_proj",
+    "transformer.h.28.mlp.w2",
+    "transformer.h.6.mlp.w2",
+    "transformer.h.13.mlp.w1",
+    "transformer.visual.transformer.resblocks.39.attn.out_proj",
+    "transformer.h.2.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.attn.out_proj",
+    "transformer.visual.transformer.resblocks.0.attn.out_proj",
+    "transformer.h.4.attn.c_proj",
+    "transformer.h.22.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.12.attn.out_proj",
+    "transformer.visual.transformer.resblocks.10.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.43.attn.in_proj",
+    "transformer.visual.transformer.resblocks.0.attn.in_proj",
+    "transformer.visual.transformer.resblocks.26.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.11.mlp.c_proj",
+    "transformer.h.0.attn.c_attn",
+    "transformer.h.19.mlp.w2",
+    "transformer.visual.transformer.resblocks.37.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.40.mlp.c_proj",
+    "transformer.h.31.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.mlp.c_fc",
+    "transformer.h.18.mlp.w1",
+    "transformer.h.23.mlp.w2",
+    "transformer.visual.transformer.resblocks.6.attn.out_proj",
+    "transformer.visual.transformer.resblocks.17.attn.in_proj",
+    "transformer.visual.transformer.resblocks.27.attn.out_proj",
+    "transformer.h.12.mlp.w2",
+    "transformer.h.23.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.29.attn.in_proj",
+    "transformer.h.10.mlp.w1",
+    "transformer.visual.transformer.resblocks.18.attn.out_proj",
+    "transformer.visual.transformer.resblocks.4.attn.out_proj",
+    "transformer.visual.transformer.resblocks.8.mlp.c_fc",
+    "transformer.h.9.mlp.w1",
+    "transformer.visual.transformer.resblocks.38.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.attn.in_proj",
+    "transformer.visual.transformer.resblocks.14.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.22.attn.in_proj",
+    "transformer.visual.transformer.resblocks.25.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.23.attn.out_proj",
+    "transformer.visual.transformer.resblocks.3.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.41.mlp.c_proj",
+    "transformer.h.24.attn.c_proj",
+    "transformer.visual.transformer.resblocks.7.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.38.mlp.c_fc",
+    "transformer.h.10.attn.c_attn",
+    "transformer.h.26.attn.c_attn",
+    "transformer.visual.transformer.resblocks.5.attn.in_proj",
+    "transformer.visual.transformer.resblocks.2.attn.out_proj",
+    "transformer.h.7.attn.c_proj",
+    "transformer.h.24.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.34.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.2.mlp.c_proj",
+    "transformer.h.12.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.14.attn.out_proj",
+    "transformer.h.18.attn.c_attn",
+    "transformer.h.23.attn.c_proj",
+    "transformer.h.27.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.26.mlp.c_proj",
+    "transformer.h.3.mlp.w1",
+    "transformer.h.2.mlp.w2",
+    "transformer.visual.transformer.resblocks.45.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.25.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.45.attn.out_proj",
+    "transformer.h.25.mlp.w1",
+    "transformer.visual.transformer.resblocks.15.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.24.attn.in_proj",
+    "transformer.h.1.attn.c_proj",
+    "transformer.h.1.attn.c_attn",
+    "transformer.visual.transformer.resblocks.4.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.9.mlp.c_fc",
+    "transformer.h.13.attn.c_attn",
+    "transformer.visual.transformer.resblocks.40.attn.out_proj",
+    "transformer.h.7.mlp.w2",
+    "transformer.h.9.attn.c_proj",
+    "transformer.h.15.attn.c_attn",
+    "transformer.visual.transformer.resblocks.0.mlp.c_fc",
+    "transformer.h.27.attn.c_attn",
+    "transformer.h.15.mlp.c_proj",
+    "transformer.h.21.mlp.w2",
+    "transformer.h.28.attn.c_proj",
+    "transformer.visual.transformer.resblocks.42.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.16.attn.out_proj",
+    "transformer.h.9.mlp.w2",
+    "transformer.visual.transformer.resblocks.9.attn.in_proj",
+    "transformer.visual.transformer.resblocks.28.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.mlp.c_proj",
+    "transformer.h.11.mlp.w1",
+    "transformer.visual.transformer.resblocks.18.attn.in_proj",
+    "transformer.h.10.attn.c_proj",
+    "transformer.visual.transformer.resblocks.42.mlp.c_fc",
+    "transformer.h.31.attn.c_attn",
+    "transformer.visual.transformer.resblocks.23.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.21.attn.in_proj",
+    "transformer.h.24.mlp.w1",
+    "transformer.visual.transformer.resblocks.35.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.7.mlp.c_proj",
+    "transformer.h.8.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.12.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.7.attn.out_proj",
+    "transformer.h.22.mlp.w2",
+    "transformer.h.29.mlp.w2",
+    "transformer.h.0.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.38.attn.in_proj",
+    "transformer.h.8.mlp.w1",
+    "transformer.h.0.mlp.w1",
+    "transformer.h.26.mlp.w2",
+    "transformer.h.25.attn.c_proj",
+    "transformer.h.27.mlp.w1",
+    "transformer.visual.transformer.resblocks.21.attn.out_proj",
+    "transformer.visual.transformer.resblocks.44.attn.in_proj",
+    "transformer.visual.transformer.resblocks.43.attn.out_proj",
+    "transformer.h.29.attn.c_attn",
+    "transformer.h.24.attn.c_attn",
+    "transformer.visual.transformer.resblocks.17.attn.out_proj",
+    "transformer.h.2.attn.c_proj",
+    "transformer.visual.transformer.resblocks.15.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.11.attn.in_proj",
+    "transformer.visual.transformer.resblocks.17.mlp.c_proj",
+    "transformer.h.11.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.41.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.5.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.18.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.24.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.attn.out_proj",
+    "transformer.h.1.mlp.w2",
+    "transformer.h.21.mlp.c_proj",
+    "transformer.h.23.attn.c_attn",
+    "transformer.visual.transformer.resblocks.34.attn.out_proj",
+    "transformer.h.14.attn.c_attn",
+    "transformer.visual.transformer.resblocks.2.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.31.attn.out_proj",
+    "transformer.visual.transformer.resblocks.30.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.11.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.31.attn.in_proj",
+    "transformer.visual.transformer.resblocks.39.mlp.c_proj",
+    "transformer.h.9.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.20.attn.out_proj",
+    "transformer.h.18.mlp.c_proj",
+    "transformer.h.19.mlp.w1",
+    "transformer.h.9.attn.c_attn",
+    "transformer.visual.transformer.resblocks.36.attn.out_proj",
+    "transformer.visual.transformer.resblocks.7.attn.in_proj",
+    "transformer.visual.transformer.resblocks.30.attn.in_proj",
+    "transformer.visual.transformer.resblocks.47.attn.out_proj",
+    "transformer.visual.transformer.resblocks.0.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.15.attn.in_proj",
+    "transformer.visual.transformer.resblocks.29.attn.out_proj",
+    "transformer.visual.transformer.resblocks.41.attn.in_proj",
+    "transformer.visual.transformer.resblocks.4.attn.in_proj",
+    "transformer.h.25.attn.c_attn",
+    "transformer.visual.transformer.resblocks.12.mlp.c_proj",
+    "transformer.h.16.mlp.w1",
+    "transformer.h.28.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.27.attn.in_proj",
+    "transformer.visual.transformer.resblocks.13.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.attn.in_proj",
+    "transformer.visual.transformer.resblocks.45.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.46.mlp.c_proj",
+    "transformer.h.30.mlp.w1",
+    "transformer.visual.transformer.resblocks.43.mlp.c_fc",
+    "transformer.h.15.mlp.w1",
+    "transformer.h.16.attn.c_proj",
+    "transformer.h.20.mlp.w1",
+    "transformer.visual.transformer.resblocks.21.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.10.mlp.c_proj",
+    "transformer.h.10.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.35.attn.in_proj",
+    "transformer.h.13.mlp.w2",
+    "transformer.visual.transformer.resblocks.8.attn.out_proj",
+    "transformer.visual.transformer.resblocks.20.mlp.c_proj",
+    "transformer.h.22.attn.c_proj",
+    "transformer.h.6.mlp.w1",
+    "transformer.h.18.mlp.w2",
+    "transformer.h.4.mlp.c_proj",
+    "transformer.h.3.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.42.attn.out_proj",
+    "transformer.visual.transformer.resblocks.36.attn.in_proj",
+    "transformer.visual.transformer.resblocks.17.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.43.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.37.attn.in_proj",
+    "transformer.visual.transformer.resblocks.1.attn.out_proj",
+    "transformer.visual.transformer.resblocks.22.mlp.c_fc",
+    "transformer.h.22.mlp.w1",
+    "transformer.visual.transformer.resblocks.44.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.37.attn.out_proj",
+    "transformer.visual.transformer.resblocks.34.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.29.mlp.c_fc",
+    "transformer.h.18.attn.c_proj",
+    "transformer.visual.transformer.resblocks.38.attn.out_proj",
+    "transformer.h.5.attn.c_attn",
+    "transformer.visual.transformer.resblocks.19.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.15.attn.out_proj",
+    "transformer.visual.transformer.resblocks.37.mlp.c_fc",
+    "transformer.h.5.attn.c_proj",
+    "transformer.h.7.attn.c_attn",
+    "transformer.visual.transformer.resblocks.28.attn.out_proj",
+    "transformer.visual.transformer.resblocks.31.mlp.c_proj",
+    "transformer.h.29.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.45.attn.in_proj",
+    "transformer.visual.transformer.resblocks.27.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.10.attn.out_proj",
+    "transformer.visual.transformer.resblocks.40.attn.in_proj",
+    "transformer.h.23.mlp.w1",
+    "transformer.visual.transformer.resblocks.28.attn.in_proj",
+    "transformer.h.12.attn.c_proj",
+    "transformer.h.16.mlp.w2",
+    "transformer.h.27.mlp.w2",
+    "transformer.visual.transformer.resblocks.22.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.9.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.47.mlp.c_proj",
+    "transformer.h.26.attn.c_proj",
+    "transformer.visual.transformer.resblocks.40.mlp.c_fc",
+    "transformer.h.8.mlp.w2",
+    "transformer.visual.transformer.resblocks.27.mlp.c_fc",
+    "transformer.h.17.mlp.w1",
+    "transformer.h.31.mlp.w2",
+    "transformer.visual.transformer.resblocks.11.attn.out_proj",
+    "transformer.h.28.mlp.w1",
+    "transformer.visual.transformer.resblocks.10.attn.in_proj",
+    "transformer.h.12.mlp.w1",
+    "transformer.h.30.mlp.w2",
+    "transformer.visual.transformer.resblocks.13.attn.in_proj",
+    "transformer.h.6.attn.c_attn",
+    "transformer.h.5.mlp.c_proj",
+    "transformer.h.6.mlp.c_proj",
+    "transformer.h.22.attn.c_attn",
+    "transformer.h.13.attn.c_proj",
+    "transformer.visual.transformer.resblocks.46.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.41.attn.out_proj",
+    "transformer.visual.transformer.resblocks.30.mlp.c_fc",
+    "transformer.h.17.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.5.attn.out_proj",
+    "transformer.h.4.mlp.w2",
+    "transformer.visual.transformer.resblocks.1.mlp.c_proj",
+    "transformer.h.11.mlp.w2",
+    "transformer.h.19.attn.c_attn",
+    "transformer.h.14.mlp.w1",
+    "transformer.visual.transformer.resblocks.44.attn.out_proj",
+    "transformer.visual.transformer.resblocks.14.mlp.c_fc",
+    "transformer.h.21.attn.c_attn",
+    "transformer.visual.transformer.resblocks.36.mlp.c_proj",
+    "transformer.h.2.mlp.w1",
+    "transformer.h.14.attn.c_proj",
+    "transformer.visual.transformer.resblocks.46.attn.in_proj",
+    "transformer.h.6.attn.c_proj",
+    "transformer.h.0.mlp.w2",
+    "transformer.h.5.mlp.w1",
+    "transformer.h.30.attn.c_proj",
+    "transformer.h.24.mlp.w2",
+    "transformer.h.0.attn.c_proj",
+    "transformer.visual.transformer.resblocks.4.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.22.attn.out_proj",
+    "transformer.h.10.mlp.w2",
+    "transformer.h.17.mlp.w2",
+    "transformer.visual.transformer.resblocks.23.attn.in_proj",
+    "transformer.visual.transformer.resblocks.36.mlp.c_fc",
+    "transformer.h.20.mlp.w2",
+    "transformer.visual.transformer.resblocks.9.attn.out_proj",
+    "transformer.h.29.mlp.w1",
+    "transformer.visual.transformer.resblocks.20.attn.in_proj",
+    "transformer.visual.transformer.resblocks.20.mlp.c_fc",
+    "transformer.h.15.attn.c_proj",
+    "transformer.h.3.mlp.w2",
+    "transformer.h.30.attn.c_attn",
+    "transformer.visual.transformer.resblocks.47.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.16.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.39.mlp.c_fc",
+    "transformer.h.20.attn.c_attn",
+    "transformer.h.19.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.46.attn.out_proj",
+    "transformer.visual.transformer.resblocks.29.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.19.attn.out_proj",
+    "transformer.visual.transformer.resblocks.26.attn.in_proj",
+    "transformer.visual.transformer.resblocks.16.mlp.c_fc",
+    "transformer.h.11.attn.c_proj",
+    "transformer.h.12.attn.c_attn",
+    "transformer.visual.conv1",
+    "transformer.visual.transformer.resblocks.35.attn.out_proj",
+    "transformer.h.25.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.14.attn.in_proj",
+    "transformer.h.26.mlp.w1",
+    "transformer.visual.transformer.resblocks.1.mlp.c_fc",
+    "transformer.h.7.mlp.c_proj",
+    "transformer.h.29.attn.c_proj",
+    "transformer.h.1.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.mlp.c_proj",
+    "transformer.h.14.mlp.c_proj",
+    "transformer.h.3.attn.c_proj",
+    "transformer.h.25.mlp.w2",
+    "transformer.h.20.attn.c_proj",
+    "transformer.h.16.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.attn.in_proj",
+    "transformer.h.17.attn.c_attn",
+    "transformer.h.14.mlp.w2",
+    "transformer.visual.transformer.resblocks.2.attn.in_proj",
+    "transformer.visual.transformer.resblocks.5.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.33.attn.out_proj",
+    "transformer.h.15.mlp.w2",
+    "transformer.h.4.attn.c_attn",
+    "transformer.h.31.mlp.w1",
+    "transformer.h.11.attn.c_attn",
+    "transformer.visual.transformer.resblocks.23.mlp.c_proj",
+    "transformer.h.7.mlp.w1",
+    "transformer.visual.transformer.resblocks.34.attn.in_proj",
+    "transformer.h.1.mlp.w1",
+    "transformer.visual.transformer.resblocks.28.mlp.c_fc",
+    "transformer.h.21.attn.c_proj",
+    "transformer.h.30.mlp.c_proj",
+    "transformer.h.21.mlp.w1",
+    "transformer.visual.transformer.resblocks.30.attn.out_proj",
+    "transformer.visual.transformer.resblocks.42.attn.in_proj",
+    "transformer.visual.transformer.resblocks.25.attn.out_proj",
+    "transformer.visual.transformer.resblocks.19.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.39.attn.in_proj",
+    "transformer.visual.transformer.resblocks.19.attn.in_proj",
+    "transformer.visual.transformer.resblocks.13.mlp.c_fc",
+    "transformer.h.13.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.25.attn.in_proj",
+    "transformer.visual.transformer.resblocks.31.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.24.attn.out_proj",
+    "transformer.visual.transformer.resblocks.24.mlp.c_fc",
+    "transformer.h.4.mlp.w1",
+    "transformer.h.8.attn.c_attn",
+    "transformer.visual.transformer.resblocks.21.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.44.mlp.c_proj",
+    "transformer.h.28.attn.c_attn",
+    "transformer.visual.transformer.resblocks.18.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.attn.in_proj",
+    "transformer.h.19.attn.c_proj",
+    "transformer.h.2.attn.c_attn",
+    "transformer.visual.transformer.resblocks.35.mlp.c_proj",
+    "transformer.h.26.mlp.c_proj",
+    "transformer.h.8.attn.c_proj",
+    "transformer.h.27.attn.c_proj",
+    "transformer.visual.transformer.resblocks.13.attn.out_proj",
+    "transformer.h.16.attn.c_attn",
+    "transformer.visual.transformer.resblocks.16.attn.in_proj",
+    "transformer.visual.transformer.resblocks.8.attn.in_proj",
+    "transformer.visual.transformer.resblocks.26.attn.out_proj",
+    "transformer.h.31.attn.c_proj",
+    "transformer.h.5.mlp.w2",
+    "transformer.visual.transformer.resblocks.8.mlp.c_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-2000/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:278f1300ea67a75f5c5a17bdd7a6059b913709dd5abe64e411b97609915f4bab
+size 469105640

checkpoint-2000/latest

@@ -0,0 +1 @@
+global_step2000

checkpoint-2000/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f6c02405ec3457460084b0bccc1f52114416050135941d1b86a40847a3901cd
+size 14960

checkpoint-2000/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8b8c688657b62198cfea2b0bfe429c988dc9d8749e2d0e57204088b7624fcfb
+size 14960

checkpoint-2000/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:836759327c6fc5baec90582cf262c9e057b66ddd65bd799ca61947470534bfd5
+size 14960

checkpoint-2000/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a63a9100ca1a6600d9304f9d2a8977a8b49d8a7a30c82ba884c0cce68472ba4b
+size 14960

checkpoint-2000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:51900721eac30baaf24efb0aa845d2e4f9b1fb9c462b5a0523edfc3c327d92c0
+size 1064

checkpoint-2000/special_tokens_map.json

@@ -0,0 +1,3 @@
+{
+  "pad_token": "<|endoftext|>"
+}

checkpoint-2000/tokenizer_config.json

@@ -0,0 +1,14 @@
+{
+  "added_tokens_decoder": {},
+  "auto_map": {
+    "AutoTokenizer": [
+      "Qwen/Qwen-VL-Chat--tokenization_qwen.QWenTokenizer",
+      null
+    ]
+  },
+  "clean_up_tokenization_spaces": true,
+  "model_max_length": 1280,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "tokenizer_class": "QWenTokenizer"
+}

checkpoint-2000/trainer_state.json

@@ -0,0 +1,1433 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.3365303718660609,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016826518593303045,
+      "grad_norm": 5.367858933563703,
+      "learning_rate": 4.9999999999999996e-06,
+      "loss": 0.9537,
+      "step": 10
+    },
+    {
+      "epoch": 0.003365303718660609,
+      "grad_norm": 9.386746384686745,
+      "learning_rate": 9.999999999999999e-06,
+      "loss": 0.943,
+      "step": 20
+    },
+    {
+      "epoch": 0.005047955577990914,
+      "grad_norm": 7.387362447577942,
+      "learning_rate": 1.5e-05,
+      "loss": 0.934,
+      "step": 30
+    },
+    {
+      "epoch": 0.006730607437321218,
+      "grad_norm": 6.9256319824932655,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 0.8376,
+      "step": 40
+    },
+    {
+      "epoch": 0.008413259296651522,
+      "grad_norm": 9.1148382590838,
+      "learning_rate": 2.5e-05,
+      "loss": 0.8484,
+      "step": 50
+    },
+    {
+      "epoch": 0.010095911155981827,
+      "grad_norm": 3.9989232759892426,
+      "learning_rate": 3e-05,
+      "loss": 0.8097,
+      "step": 60
+    },
+    {
+      "epoch": 0.011778563015312132,
+      "grad_norm": 3.892371218590039,
+      "learning_rate": 2.9999786123888308e-05,
+      "loss": 0.7811,
+      "step": 70
+    },
+    {
+      "epoch": 0.013461214874642436,
+      "grad_norm": 8.096662196282066,
+      "learning_rate": 2.9999144501652298e-05,
+      "loss": 0.7446,
+      "step": 80
+    },
+    {
+      "epoch": 0.01514386673397274,
+      "grad_norm": 1.5769306611206149,
+      "learning_rate": 2.9998075151588992e-05,
+      "loss": 0.7258,
+      "step": 90
+    },
+    {
+      "epoch": 0.016826518593303044,
+      "grad_norm": 8.47430485487969,
+      "learning_rate": 2.999657810419285e-05,
+      "loss": 0.7052,
+      "step": 100
+    },
+    {
+      "epoch": 0.01850917045263335,
+      "grad_norm": 2.363071299913598,
+      "learning_rate": 2.999465340215489e-05,
+      "loss": 0.7657,
+      "step": 110
+    },
+    {
+      "epoch": 0.020191822311963654,
+      "grad_norm": 1.9252385425154874,
+      "learning_rate": 2.999230110036149e-05,
+      "loss": 0.7329,
+      "step": 120
+    },
+    {
+      "epoch": 0.02187447417129396,
+      "grad_norm": 8.946028475031488,
+      "learning_rate": 2.99895212658928e-05,
+      "loss": 0.7304,
+      "step": 130
+    },
+    {
+      "epoch": 0.023557126030624265,
+      "grad_norm": 6.877609312630206,
+      "learning_rate": 2.9986313978020846e-05,
+      "loss": 0.7453,
+      "step": 140
+    },
+    {
+      "epoch": 0.02523977788995457,
+      "grad_norm": 2.5256324882367993,
+      "learning_rate": 2.9982679328207262e-05,
+      "loss": 0.7366,
+      "step": 150
+    },
+    {
+      "epoch": 0.02692242974928487,
+      "grad_norm": 2.709550398238738,
+      "learning_rate": 2.9978617420100692e-05,
+      "loss": 0.7258,
+      "step": 160
+    },
+    {
+      "epoch": 0.028605081608615177,
+      "grad_norm": 1.543550019689774,
+      "learning_rate": 2.9974128369533805e-05,
+      "loss": 0.7372,
+      "step": 170
+    },
+    {
+      "epoch": 0.03028773346794548,
+      "grad_norm": 3.3453966881155504,
+      "learning_rate": 2.9969212304520034e-05,
+      "loss": 0.743,
+      "step": 180
+    },
+    {
+      "epoch": 0.03197038532727579,
+      "grad_norm": 1.922001656181265,
+      "learning_rate": 2.9963869365249895e-05,
+      "loss": 0.7819,
+      "step": 190
+    },
+    {
+      "epoch": 0.03365303718660609,
+      "grad_norm": 2.0611188483400036,
+      "learning_rate": 2.995809970408699e-05,
+      "loss": 0.7155,
+      "step": 200
+    },
+    {
+      "epoch": 0.0353356890459364,
+      "grad_norm": 1.5313041833127259,
+      "learning_rate": 2.9951903485563685e-05,
+      "loss": 0.7322,
+      "step": 210
+    },
+    {
+      "epoch": 0.0370183409052667,
+      "grad_norm": 2.0124191694435085,
+      "learning_rate": 2.99452808863764e-05,
+      "loss": 0.6759,
+      "step": 220
+    },
+    {
+      "epoch": 0.03870099276459701,
+      "grad_norm": 3.182123324389477,
+      "learning_rate": 2.993823209538056e-05,
+      "loss": 0.6953,
+      "step": 230
+    },
+    {
+      "epoch": 0.04038364462392731,
+      "grad_norm": 1.6122782177661379,
+      "learning_rate": 2.9930757313585238e-05,
+      "loss": 0.6953,
+      "step": 240
+    },
+    {
+      "epoch": 0.04206629648325761,
+      "grad_norm": 2.2027482596695647,
+      "learning_rate": 2.9922856754147406e-05,
+      "loss": 0.7301,
+      "step": 250
+    },
+    {
+      "epoch": 0.04374894834258792,
+      "grad_norm": 2.6782477155989213,
+      "learning_rate": 2.9914530642365852e-05,
+      "loss": 0.6891,
+      "step": 260
+    },
+    {
+      "epoch": 0.04543160020191822,
+      "grad_norm": 1.9740401144541417,
+      "learning_rate": 2.990577921567476e-05,
+      "loss": 0.7231,
+      "step": 270
+    },
+    {
+      "epoch": 0.04711425206124853,
+      "grad_norm": 1.719874620968932,
+      "learning_rate": 2.989660272363696e-05,
+      "loss": 0.7505,
+      "step": 280
+    },
+    {
+      "epoch": 0.04879690392057883,
+      "grad_norm": 1.3138364164203409,
+      "learning_rate": 2.988700142793676e-05,
+      "loss": 0.7116,
+      "step": 290
+    },
+    {
+      "epoch": 0.05047955577990914,
+      "grad_norm": 5.853627389344256,
+      "learning_rate": 2.9876975602372536e-05,
+      "loss": 0.719,
+      "step": 300
+    },
+    {
+      "epoch": 0.05216220763923944,
+      "grad_norm": 2.347259437170711,
+      "learning_rate": 2.9866525532848906e-05,
+      "loss": 0.6803,
+      "step": 310
+    },
+    {
+      "epoch": 0.05384485949856974,
+      "grad_norm": 1.937679220955038,
+      "learning_rate": 2.9855651517368567e-05,
+      "loss": 0.7461,
+      "step": 320
+    },
+    {
+      "epoch": 0.05552751135790005,
+      "grad_norm": 1.6661300351569575,
+      "learning_rate": 2.9844353866023802e-05,
+      "loss": 0.7472,
+      "step": 330
+    },
+    {
+      "epoch": 0.05721016321723035,
+      "grad_norm": 2.357915869204484,
+      "learning_rate": 2.9832632900987642e-05,
+      "loss": 0.7148,
+      "step": 340
+    },
+    {
+      "epoch": 0.05889281507656066,
+      "grad_norm": 4.398815186243292,
+      "learning_rate": 2.982048895650468e-05,
+      "loss": 0.6992,
+      "step": 350
+    },
+    {
+      "epoch": 0.06057546693589096,
+      "grad_norm": 12.662682224480092,
+      "learning_rate": 2.9807922378881537e-05,
+      "loss": 0.7539,
+      "step": 360
+    },
+    {
+      "epoch": 0.06225811879522127,
+      "grad_norm": 0.8642696401357872,
+      "learning_rate": 2.979493352647697e-05,
+      "loss": 0.7212,
+      "step": 370
+    },
+    {
+      "epoch": 0.06394077065455157,
+      "grad_norm": 27.047937858232604,
+      "learning_rate": 2.9781522769691686e-05,
+      "loss": 0.722,
+      "step": 380
+    },
+    {
+      "epoch": 0.06562342251388188,
+      "grad_norm": 2.598805292448644,
+      "learning_rate": 2.9767690490957758e-05,
+      "loss": 0.7065,
+      "step": 390
+    },
+    {
+      "epoch": 0.06730607437321218,
+      "grad_norm": 1.2314762895092763,
+      "learning_rate": 2.9753437084727713e-05,
+      "loss": 0.7498,
+      "step": 400
+    },
+    {
+      "epoch": 0.06898872623254249,
+      "grad_norm": 1.6421909669790502,
+      "learning_rate": 2.9738762957463292e-05,
+      "loss": 0.6992,
+      "step": 410
+    },
+    {
+      "epoch": 0.0706713780918728,
+      "grad_norm": 2.023552968622588,
+      "learning_rate": 2.9723668527623877e-05,
+      "loss": 0.6943,
+      "step": 420
+    },
+    {
+      "epoch": 0.0723540299512031,
+      "grad_norm": 1.5172337910969138,
+      "learning_rate": 2.9708154225654526e-05,
+      "loss": 0.6987,
+      "step": 430
+    },
+    {
+      "epoch": 0.0740366818105334,
+      "grad_norm": 1.197852135730745,
+      "learning_rate": 2.9692220493973712e-05,
+      "loss": 0.7302,
+      "step": 440
+    },
+    {
+      "epoch": 0.0757193336698637,
+      "grad_norm": 2.4396443837967183,
+      "learning_rate": 2.9675867786960718e-05,
+      "loss": 0.7318,
+      "step": 450
+    },
+    {
+      "epoch": 0.07740198552919401,
+      "grad_norm": 1.4599851880563282,
+      "learning_rate": 2.9659096570942654e-05,
+      "loss": 0.6941,
+      "step": 460
+    },
+    {
+      "epoch": 0.07908463738852431,
+      "grad_norm": 1.117755825364562,
+      "learning_rate": 2.9641907324181194e-05,
+      "loss": 0.7399,
+      "step": 470
+    },
+    {
+      "epoch": 0.08076728924785462,
+      "grad_norm": 2.9235378164576242,
+      "learning_rate": 2.96243005368589e-05,
+      "loss": 0.7207,
+      "step": 480
+    },
+    {
+      "epoch": 0.08244994110718493,
+      "grad_norm": 7.308883163781362,
+      "learning_rate": 2.960627671106527e-05,
+      "loss": 0.682,
+      "step": 490
+    },
+    {
+      "epoch": 0.08413259296651522,
+      "grad_norm": 3.4394827932955234,
+      "learning_rate": 2.9587836360782405e-05,
+      "loss": 0.708,
+      "step": 500
+    },
+    {
+      "epoch": 0.08581524482584553,
+      "grad_norm": 3.2314529856927634,
+      "learning_rate": 2.9568980011870357e-05,
+      "loss": 0.7335,
+      "step": 510
+    },
+    {
+      "epoch": 0.08749789668517584,
+      "grad_norm": 1.825724533695325,
+      "learning_rate": 2.954970820205214e-05,
+      "loss": 0.6951,
+      "step": 520
+    },
+    {
+      "epoch": 0.08918054854450615,
+      "grad_norm": 3.3231741746640076,
+      "learning_rate": 2.9530021480898393e-05,
+      "loss": 0.7793,
+      "step": 530
+    },
+    {
+      "epoch": 0.09086320040383644,
+      "grad_norm": 1.3097651462571123,
+      "learning_rate": 2.9509920409811696e-05,
+      "loss": 0.7087,
+      "step": 540
+    },
+    {
+      "epoch": 0.09254585226316675,
+      "grad_norm": 6.685911471215255,
+      "learning_rate": 2.9489405562010565e-05,
+      "loss": 0.6906,
+      "step": 550
+    },
+    {
+      "epoch": 0.09422850412249706,
+      "grad_norm": 2.870746617513948,
+      "learning_rate": 2.9468477522513132e-05,
+      "loss": 0.7028,
+      "step": 560
+    },
+    {
+      "epoch": 0.09591115598182735,
+      "grad_norm": 1.782555352805469,
+      "learning_rate": 2.9447136888120408e-05,
+      "loss": 0.6901,
+      "step": 570
+    },
+    {
+      "epoch": 0.09759380784115766,
+      "grad_norm": 2.336519711000487,
+      "learning_rate": 2.9425384267399327e-05,
+      "loss": 0.7779,
+      "step": 580
+    },
+    {
+      "epoch": 0.09927645970048797,
+      "grad_norm": 8.935574410818228,
+      "learning_rate": 2.940322028066534e-05,
+      "loss": 0.7503,
+      "step": 590
+    },
+    {
+      "epoch": 0.10095911155981828,
+      "grad_norm": 2.754713786882031,
+      "learning_rate": 2.938064555996476e-05,
+      "loss": 0.7208,
+      "step": 600
+    },
+    {
+      "epoch": 0.10264176341914857,
+      "grad_norm": 1.5082503557652136,
+      "learning_rate": 2.9357660749056713e-05,
+      "loss": 0.7169,
+      "step": 610
+    },
+    {
+      "epoch": 0.10432441527847888,
+      "grad_norm": 9.04522194526273,
+      "learning_rate": 2.9334266503394803e-05,
+      "loss": 0.6927,
+      "step": 620
+    },
+    {
+      "epoch": 0.10600706713780919,
+      "grad_norm": 55.28278686388287,
+      "learning_rate": 2.9310463490108397e-05,
+      "loss": 0.7107,
+      "step": 630
+    },
+    {
+      "epoch": 0.10768971899713949,
+      "grad_norm": 3.721916069105249,
+      "learning_rate": 2.928625238798362e-05,
+      "loss": 0.6951,
+      "step": 640
+    },
+    {
+      "epoch": 0.1093723708564698,
+      "grad_norm": 2.5040797323750112,
+      "learning_rate": 2.9261633887443993e-05,
+      "loss": 0.6916,
+      "step": 650
+    },
+    {
+      "epoch": 0.1110550227158001,
+      "grad_norm": 3.5468924769840617,
+      "learning_rate": 2.9236608690530738e-05,
+      "loss": 0.7077,
+      "step": 660
+    },
+    {
+      "epoch": 0.11273767457513041,
+      "grad_norm": 3.0266819778200746,
+      "learning_rate": 2.921117751088276e-05,
+      "loss": 0.6952,
+      "step": 670
+    },
+    {
+      "epoch": 0.1144203264344607,
+      "grad_norm": 1.634743894298146,
+      "learning_rate": 2.91853410737163e-05,
+      "loss": 0.6936,
+      "step": 680
+    },
+    {
+      "epoch": 0.11610297829379101,
+      "grad_norm": 1.0925365801520501,
+      "learning_rate": 2.915910011580426e-05,
+      "loss": 0.7317,
+      "step": 690
+    },
+    {
+      "epoch": 0.11778563015312132,
+      "grad_norm": 1.6959112138540386,
+      "learning_rate": 2.9132455385455176e-05,
+      "loss": 0.6917,
+      "step": 700
+    },
+    {
+      "epoch": 0.11946828201245162,
+      "grad_norm": 1.9723433746891168,
+      "learning_rate": 2.9105407642491895e-05,
+      "loss": 0.7209,
+      "step": 710
+    },
+    {
+      "epoch": 0.12115093387178193,
+      "grad_norm": 2.1537215293733833,
+      "learning_rate": 2.907795765822989e-05,
+      "loss": 0.7488,
+      "step": 720
+    },
+    {
+      "epoch": 0.12283358573111224,
+      "grad_norm": 3.227101869737169,
+      "learning_rate": 2.9050106215455283e-05,
+      "loss": 0.7152,
+      "step": 730
+    },
+    {
+      "epoch": 0.12451623759044254,
+      "grad_norm": 2.7222358893572554,
+      "learning_rate": 2.9021854108402516e-05,
+      "loss": 0.708,
+      "step": 740
+    },
+    {
+      "epoch": 0.12619888944977284,
+      "grad_norm": 2.1054843767538136,
+      "learning_rate": 2.8993202142731693e-05,
+      "loss": 0.7251,
+      "step": 750
+    },
+    {
+      "epoch": 0.12788154130910315,
+      "grad_norm": 2.11845883419618,
+      "learning_rate": 2.8964151135505616e-05,
+      "loss": 0.7405,
+      "step": 760
+    },
+    {
+      "epoch": 0.12956419316843346,
+      "grad_norm": 13.171512404187755,
+      "learning_rate": 2.8934701915166477e-05,
+      "loss": 0.6844,
+      "step": 770
+    },
+    {
+      "epoch": 0.13124684502776376,
+      "grad_norm": 2.7633375632879127,
+      "learning_rate": 2.890485532151225e-05,
+      "loss": 0.6766,
+      "step": 780
+    },
+    {
+      "epoch": 0.13292949688709407,
+      "grad_norm": 1.8420785342693768,
+      "learning_rate": 2.887461220567271e-05,
+      "loss": 0.7037,
+      "step": 790
+    },
+    {
+      "epoch": 0.13461214874642435,
+      "grad_norm": 1.5557447509529954,
+      "learning_rate": 2.8843973430085204e-05,
+      "loss": 0.6991,
+      "step": 800
+    },
+    {
+      "epoch": 0.13629480060575466,
+      "grad_norm": 1.9295826624758823,
+      "learning_rate": 2.8812939868470016e-05,
+      "loss": 0.6956,
+      "step": 810
+    },
+    {
+      "epoch": 0.13797745246508497,
+      "grad_norm": 3.3211216557707126,
+      "learning_rate": 2.878151240580548e-05,
+      "loss": 0.6774,
+      "step": 820
+    },
+    {
+      "epoch": 0.13966010432441528,
+      "grad_norm": 4.196064403930616,
+      "learning_rate": 2.874969193830274e-05,
+      "loss": 0.6752,
+      "step": 830
+    },
+    {
+      "epoch": 0.1413427561837456,
+      "grad_norm": 5.574976270137628,
+      "learning_rate": 2.871747937338016e-05,
+      "loss": 0.6553,
+      "step": 840
+    },
+    {
+      "epoch": 0.1430254080430759,
+      "grad_norm": 1.6494038718740478,
+      "learning_rate": 2.8684875629637505e-05,
+      "loss": 0.7152,
+      "step": 850
+    },
+    {
+      "epoch": 0.1447080599024062,
+      "grad_norm": 1.3061892609414858,
+      "learning_rate": 2.8651881636829698e-05,
+      "loss": 0.7462,
+      "step": 860
+    },
+    {
+      "epoch": 0.1463907117617365,
+      "grad_norm": 4.321044418392694,
+      "learning_rate": 2.861849833584032e-05,
+      "loss": 0.6902,
+      "step": 870
+    },
+    {
+      "epoch": 0.1480733636210668,
+      "grad_norm": 2.9444722968009764,
+      "learning_rate": 2.8584726678654787e-05,
+      "loss": 0.6813,
+      "step": 880
+    },
+    {
+      "epoch": 0.1497560154803971,
+      "grad_norm": 1.4940245340163587,
+      "learning_rate": 2.85505676283332e-05,
+      "loss": 0.689,
+      "step": 890
+    },
+    {
+      "epoch": 0.1514386673397274,
+      "grad_norm": 3.3704010040589565,
+      "learning_rate": 2.851602215898287e-05,
+      "loss": 0.6953,
+      "step": 900
+    },
+    {
+      "epoch": 0.15312131919905772,
+      "grad_norm": 1.6597144402924948,
+      "learning_rate": 2.8481091255730552e-05,
+      "loss": 0.7277,
+      "step": 910
+    },
+    {
+      "epoch": 0.15480397105838803,
+      "grad_norm": 10.969872224353953,
+      "learning_rate": 2.844577591469435e-05,
+      "loss": 0.7142,
+      "step": 920
+    },
+    {
+      "epoch": 0.15648662291771834,
+      "grad_norm": 8.45616831264245,
+      "learning_rate": 2.8410077142955304e-05,
+      "loss": 0.7197,
+      "step": 930
+    },
+    {
+      "epoch": 0.15816927477704862,
+      "grad_norm": 2.9594258901214427,
+      "learning_rate": 2.8373995958528683e-05,
+      "loss": 0.7351,
+      "step": 940
+    },
+    {
+      "epoch": 0.15985192663637893,
+      "grad_norm": 2.168676312428759,
+      "learning_rate": 2.8337533390334942e-05,
+      "loss": 0.7544,
+      "step": 950
+    },
+    {
+      "epoch": 0.16153457849570924,
+      "grad_norm": 7.898767360662744,
+      "learning_rate": 2.8300690478170388e-05,
+      "loss": 0.7015,
+      "step": 960
+    },
+    {
+      "epoch": 0.16321723035503954,
+      "grad_norm": 16.83650212945308,
+      "learning_rate": 2.826346827267753e-05,
+      "loss": 0.7139,
+      "step": 970
+    },
+    {
+      "epoch": 0.16489988221436985,
+      "grad_norm": 2.3791337429068977,
+      "learning_rate": 2.8225867835315114e-05,
+      "loss": 0.7053,
+      "step": 980
+    },
+    {
+      "epoch": 0.16658253407370016,
+      "grad_norm": 1.9679363325295285,
+      "learning_rate": 2.8187890238327842e-05,
+      "loss": 0.7313,
+      "step": 990
+    },
+    {
+      "epoch": 0.16826518593303044,
+      "grad_norm": 1.4822625638777076,
+      "learning_rate": 2.814953656471583e-05,
+      "loss": 0.7085,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16994783779236075,
+      "grad_norm": 2.647291447509443,
+      "learning_rate": 2.8110807908203682e-05,
+      "loss": 0.6638,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17163048965169106,
+      "grad_norm": 2.969379719654364,
+      "learning_rate": 2.8071705373209328e-05,
+      "loss": 0.6884,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17331314151102137,
+      "grad_norm": 1.1163745403124403,
+      "learning_rate": 2.803223007481252e-05,
+      "loss": 0.6885,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17499579337035168,
+      "grad_norm": 1.2686557979094786,
+      "learning_rate": 2.7992383138723034e-05,
+      "loss": 0.7037,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17667844522968199,
+      "grad_norm": 4.648945448875594,
+      "learning_rate": 2.7952165701248573e-05,
+      "loss": 0.6933,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1783610970890123,
+      "grad_norm": 4.723564874595428,
+      "learning_rate": 2.7911578909262353e-05,
+      "loss": 0.7144,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18004374894834257,
+      "grad_norm": 5.211806926801946,
+      "learning_rate": 2.787062392017041e-05,
+      "loss": 0.7266,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18172640080767288,
+      "grad_norm": 1.3725560316172503,
+      "learning_rate": 2.7829301901878592e-05,
+      "loss": 0.7445,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1834090526670032,
+      "grad_norm": 0.9012241436004484,
+      "learning_rate": 2.7787614032759243e-05,
+      "loss": 0.6986,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1850917045263335,
+      "grad_norm": 2.912544243603394,
+      "learning_rate": 2.7745561501617605e-05,
+      "loss": 0.7173,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1867743563856638,
+      "grad_norm": 1.4248442614931247,
+      "learning_rate": 2.7703145507657923e-05,
+      "loss": 0.7035,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18845700824499412,
+      "grad_norm": 2.186609904533333,
+      "learning_rate": 2.766036726044926e-05,
+      "loss": 0.7371,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19013966010432443,
+      "grad_norm": 2.0524595532166603,
+      "learning_rate": 2.7617227979890957e-05,
+      "loss": 0.6986,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1918223119636547,
+      "grad_norm": 1.8227045280907195,
+      "learning_rate": 2.7573728896177897e-05,
+      "loss": 0.7075,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19350496382298502,
+      "grad_norm": 1.8425998009576734,
+      "learning_rate": 2.7529871249765397e-05,
+      "loss": 0.6897,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19518761568231532,
+      "grad_norm": 5.3035191638420836,
+      "learning_rate": 2.7485656291333845e-05,
+      "loss": 0.7027,
+      "step": 1160
+    },
+    {
+      "epoch": 0.19687026754164563,
+      "grad_norm": 3.3228474353685504,
+      "learning_rate": 2.7441085281753028e-05,
+      "loss": 0.7091,
+      "step": 1170
+    },
+    {
+      "epoch": 0.19855291940097594,
+      "grad_norm": 3.5016968564731283,
+      "learning_rate": 2.739615949204617e-05,
+      "loss": 0.7241,
+      "step": 1180
+    },
+    {
+      "epoch": 0.20023557126030625,
+      "grad_norm": 1.7190048028902127,
+      "learning_rate": 2.7350880203353703e-05,
+      "loss": 0.7192,
+      "step": 1190
+    },
+    {
+      "epoch": 0.20191822311963656,
+      "grad_norm": 3.7186824247487515,
+      "learning_rate": 2.7305248706896722e-05,
+      "loss": 0.7063,
+      "step": 1200
+    },
+    {
+      "epoch": 0.20360087497896684,
+      "grad_norm": 4.1717869895766935,
+      "learning_rate": 2.7259266303940164e-05,
+      "loss": 0.7088,
+      "step": 1210
+    },
+    {
+      "epoch": 0.20528352683829715,
+      "grad_norm": 2.5124857963805804,
+      "learning_rate": 2.7212934305755697e-05,
+      "loss": 0.7198,
+      "step": 1220
+    },
+    {
+      "epoch": 0.20696617869762746,
+      "grad_norm": 2.095136268936366,
+      "learning_rate": 2.7166254033584343e-05,
+      "loss": 0.753,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20864883055695777,
+      "grad_norm": 3.2661098868577256,
+      "learning_rate": 2.7119226818598784e-05,
+      "loss": 0.6779,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21033148241628807,
+      "grad_norm": 3.055506603735091,
+      "learning_rate": 2.7071854001865402e-05,
+      "loss": 0.7013,
+      "step": 1250
+    },
+    {
+      "epoch": 0.21201413427561838,
+      "grad_norm": 12.522953778477769,
+      "learning_rate": 2.702413693430604e-05,
+      "loss": 0.7088,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2136967861349487,
+      "grad_norm": 3.476240301739368,
+      "learning_rate": 2.697607697665948e-05,
+      "loss": 0.689,
+      "step": 1270
+    },
+    {
+      "epoch": 0.21537943799427897,
+      "grad_norm": 1.1862686197570156,
+      "learning_rate": 2.6927675499442648e-05,
+      "loss": 0.7243,
+      "step": 1280
+    },
+    {
+      "epoch": 0.21706208985360928,
+      "grad_norm": 1.6505042403801382,
+      "learning_rate": 2.68789338829115e-05,
+      "loss": 0.7083,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2187447417129396,
+      "grad_norm": 4.74071740077375,
+      "learning_rate": 2.6829853517021698e-05,
+      "loss": 0.7016,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2204273935722699,
+      "grad_norm": 4.124079283639458,
+      "learning_rate": 2.6780435801388945e-05,
+      "loss": 0.7077,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2221100454316002,
+      "grad_norm": 1.9487864410536297,
+      "learning_rate": 2.6730682145249093e-05,
+      "loss": 0.7355,
+      "step": 1320
+    },
+    {
+      "epoch": 0.22379269729093051,
+      "grad_norm": 2.4839241050514733,
+      "learning_rate": 2.668059396741795e-05,
+      "loss": 0.7092,
+      "step": 1330
+    },
+    {
+      "epoch": 0.22547534915026082,
+      "grad_norm": 2.841913657394254,
+      "learning_rate": 2.6630172696250804e-05,
+      "loss": 0.7303,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2271580010095911,
+      "grad_norm": 2.7442870185873347,
+      "learning_rate": 2.6579419769601715e-05,
+      "loss": 0.6739,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2288406528689214,
+      "grad_norm": 1.3854365909071105,
+      "learning_rate": 2.6528336634782493e-05,
+      "loss": 0.7073,
+      "step": 1360
+    },
+    {
+      "epoch": 0.23052330472825172,
+      "grad_norm": 3.115941001607779,
+      "learning_rate": 2.6476924748521443e-05,
+      "loss": 0.7267,
+      "step": 1370
+    },
+    {
+      "epoch": 0.23220595658758203,
+      "grad_norm": 6.9185951332741,
+      "learning_rate": 2.6425185576921812e-05,
+      "loss": 0.7456,
+      "step": 1380
+    },
+    {
+      "epoch": 0.23388860844691234,
+      "grad_norm": 2.378601355345996,
+      "learning_rate": 2.637312059541997e-05,
+      "loss": 0.6912,
+      "step": 1390
+    },
+    {
+      "epoch": 0.23557126030624265,
+      "grad_norm": 2.7929947858543906,
+      "learning_rate": 2.632073128874336e-05,
+      "loss": 0.7184,
+      "step": 1400
+    },
+    {
+      "epoch": 0.23725391216557296,
+      "grad_norm": 1.5382855773213957,
+      "learning_rate": 2.6268019150868144e-05,
+      "loss": 0.7099,
+      "step": 1410
+    },
+    {
+      "epoch": 0.23893656402490324,
+      "grad_norm": 6.1010563795570025,
+      "learning_rate": 2.62149856849766e-05,
+      "loss": 0.6895,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24061921588423354,
+      "grad_norm": 5.999491987974443,
+      "learning_rate": 2.616163240341426e-05,
+      "loss": 0.7493,
+      "step": 1430
+    },
+    {
+      "epoch": 0.24230186774356385,
+      "grad_norm": 2.837037600849311,
+      "learning_rate": 2.6107960827646774e-05,
+      "loss": 0.7176,
+      "step": 1440
+    },
+    {
+      "epoch": 0.24398451960289416,
+      "grad_norm": 1.7029089834427125,
+      "learning_rate": 2.6053972488216538e-05,
+      "loss": 0.6852,
+      "step": 1450
+    },
+    {
+      "epoch": 0.24566717146222447,
+      "grad_norm": 1.382189249222589,
+      "learning_rate": 2.5999668924699035e-05,
+      "loss": 0.685,
+      "step": 1460
+    },
+    {
+      "epoch": 0.24734982332155478,
+      "grad_norm": 1.9496045543050813,
+      "learning_rate": 2.5945051685658923e-05,
+      "loss": 0.6591,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2490324751808851,
+      "grad_norm": 5.479390805764353,
+      "learning_rate": 2.5890122328605908e-05,
+      "loss": 0.7085,
+      "step": 1480
+    },
+    {
+      "epoch": 0.25071512704021537,
+      "grad_norm": 1.7567995670915637,
+      "learning_rate": 2.5834882419950295e-05,
+      "loss": 0.7091,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2523977788995457,
+      "grad_norm": 1.9685911084195309,
+      "learning_rate": 2.577933353495833e-05,
+      "loss": 0.7218,
+      "step": 1500
+    },
+    {
+      "epoch": 0.254080430758876,
+      "grad_norm": 3.400633915540874,
+      "learning_rate": 2.5723477257707293e-05,
+      "loss": 0.7148,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2557630826182063,
+      "grad_norm": 1.2116738326443663,
+      "learning_rate": 2.566731518104029e-05,
+      "loss": 0.7321,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2574457344775366,
+      "grad_norm": 1.3376343864594256,
+      "learning_rate": 2.5610848906520878e-05,
+      "loss": 0.748,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2591283863368669,
+      "grad_norm": 2.6089861003232055,
+      "learning_rate": 2.5554080044387344e-05,
+      "loss": 0.7127,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2608110381961972,
+      "grad_norm": 3.2047926120640526,
+      "learning_rate": 2.5497010213506825e-05,
+      "loss": 0.7262,
+      "step": 1550
+    },
+    {
+      "epoch": 0.26249369005552753,
+      "grad_norm": 1.4899957348295265,
+      "learning_rate": 2.5439641041329128e-05,
+      "loss": 0.7122,
+      "step": 1560
+    },
+    {
+      "epoch": 0.26417634191485784,
+      "grad_norm": 3.595968473922136,
+      "learning_rate": 2.5381974163840313e-05,
+      "loss": 0.7092,
+      "step": 1570
+    },
+    {
+      "epoch": 0.26585899377418815,
+      "grad_norm": 3.5232117574234003,
+      "learning_rate": 2.532401122551605e-05,
+      "loss": 0.6924,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2675416456335184,
+      "grad_norm": 2.618947453668302,
+      "learning_rate": 2.526575387927473e-05,
+      "loss": 0.7067,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2692242974928487,
+      "grad_norm": 3.6282673284589566,
+      "learning_rate": 2.52072037864303e-05,
+      "loss": 0.6945,
+      "step": 1600
+    },
+    {
+      "epoch": 0.270906949352179,
+      "grad_norm": 2.2274379147013,
+      "learning_rate": 2.5148362616644926e-05,
+      "loss": 0.6727,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2725896012115093,
+      "grad_norm": 2.823867881580523,
+      "learning_rate": 2.508923204788135e-05,
+      "loss": 0.7158,
+      "step": 1620
+    },
+    {
+      "epoch": 0.27427225307083963,
+      "grad_norm": 2.0118901151982245,
+      "learning_rate": 2.5029813766355062e-05,
+      "loss": 0.7422,
+      "step": 1630
+    },
+    {
+      "epoch": 0.27595490493016994,
+      "grad_norm": 1.2843584175617246,
+      "learning_rate": 2.4970109466486202e-05,
+      "loss": 0.7099,
+      "step": 1640
+    },
+    {
+      "epoch": 0.27763755678950025,
+      "grad_norm": 3.5059277881120914,
+      "learning_rate": 2.491012085085122e-05,
+      "loss": 0.7164,
+      "step": 1650
+    },
+    {
+      "epoch": 0.27932020864883056,
+      "grad_norm": 1.7458993688338285,
+      "learning_rate": 2.4849849630134384e-05,
+      "loss": 0.6901,
+      "step": 1660
+    },
+    {
+      "epoch": 0.28100286050816087,
+      "grad_norm": 5.813346226937464,
+      "learning_rate": 2.4789297523078924e-05,
+      "loss": 0.7181,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2826855123674912,
+      "grad_norm": 2.0515286491489237,
+      "learning_rate": 2.4728466256438072e-05,
+      "loss": 0.7431,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2843681642268215,
+      "grad_norm": 2.6702746679350375,
+      "learning_rate": 2.4667357564925798e-05,
+      "loss": 0.701,
+      "step": 1690
+    },
+    {
+      "epoch": 0.2860508160861518,
+      "grad_norm": 2.707565805299449,
+      "learning_rate": 2.460597319116735e-05,
+      "loss": 0.6725,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2877334679454821,
+      "grad_norm": 1.7994267796032153,
+      "learning_rate": 2.4544314885649552e-05,
+      "loss": 0.7043,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2894161198048124,
+      "grad_norm": 2.240627477157692,
+      "learning_rate": 2.4482384406670883e-05,
+      "loss": 0.7337,
+      "step": 1720
+    },
+    {
+      "epoch": 0.29109877166414266,
+      "grad_norm": 1.4093208691675285,
+      "learning_rate": 2.4420183520291354e-05,
+      "loss": 0.706,
+      "step": 1730
+    },
+    {
+      "epoch": 0.292781423523473,
+      "grad_norm": 1.5799653304195502,
+      "learning_rate": 2.4357714000282127e-05,
+      "loss": 0.7254,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2944640753828033,
+      "grad_norm": 1.8282839714116759,
+      "learning_rate": 2.4294977628074938e-05,
+      "loss": 0.68,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2961467272421336,
+      "grad_norm": 13.490769798309381,
+      "learning_rate": 2.42319761927113e-05,
+      "loss": 0.6984,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2978293791014639,
+      "grad_norm": 1.1660842236351188,
+      "learning_rate": 2.4168711490791484e-05,
+      "loss": 0.6893,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2995120309607942,
+      "grad_norm": 1.4880113732457052,
+      "learning_rate": 2.4105185326423286e-05,
+      "loss": 0.7371,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3011946828201245,
+      "grad_norm": 1.9796491202207207,
+      "learning_rate": 2.4041399511170574e-05,
+      "loss": 0.7372,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3028773346794548,
+      "grad_norm": 3.2861914347482846,
+      "learning_rate": 2.3977355864001635e-05,
+      "loss": 0.7145,
+      "step": 1800
+    },
+    {
+      "epoch": 0.30455998653878513,
+      "grad_norm": 3.8536888582450595,
+      "learning_rate": 2.3913056211237304e-05,
+      "loss": 0.7244,
+      "step": 1810
+    },
+    {
+      "epoch": 0.30624263839811544,
+      "grad_norm": 2.250827213388724,
+      "learning_rate": 2.3848502386498866e-05,
+      "loss": 0.7444,
+      "step": 1820
+    },
+    {
+      "epoch": 0.30792529025744575,
+      "grad_norm": 1.6760548188250846,
+      "learning_rate": 2.3783696230655802e-05,
+      "loss": 0.7415,
+      "step": 1830
+    },
+    {
+      "epoch": 0.30960794211677606,
+      "grad_norm": 2.83690011157284,
+      "learning_rate": 2.371863959177326e-05,
+      "loss": 0.6769,
+      "step": 1840
+    },
+    {
+      "epoch": 0.31129059397610637,
+      "grad_norm": 3.6586666108883037,
+      "learning_rate": 2.365333432505937e-05,
+      "loss": 0.6981,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3129732458354367,
+      "grad_norm": 2.967916913846329,
+      "learning_rate": 2.3587782292812323e-05,
+      "loss": 0.7235,
+      "step": 1860
+    },
+    {
+      "epoch": 0.31465589769476693,
+      "grad_norm": 2.7607388194454607,
+      "learning_rate": 2.35219853643673e-05,
+      "loss": 0.7202,
+      "step": 1870
+    },
+    {
+      "epoch": 0.31633854955409724,
+      "grad_norm": 2.5793375573884925,
+      "learning_rate": 2.3455945416043132e-05,
+      "loss": 0.7437,
+      "step": 1880
+    },
+    {
+      "epoch": 0.31802120141342755,
+      "grad_norm": 1.6474727320404343,
+      "learning_rate": 2.338966433108879e-05,
+      "loss": 0.6664,
+      "step": 1890
+    },
+    {
+      "epoch": 0.31970385327275785,
+      "grad_norm": 2.8252072958720102,
+      "learning_rate": 2.3323143999629712e-05,
+      "loss": 0.6641,
+      "step": 1900
+    },
+    {
+      "epoch": 0.32138650513208816,
+      "grad_norm": 1.8240997471681801,
+      "learning_rate": 2.3256386318613877e-05,
+      "loss": 0.7029,
+      "step": 1910
+    },
+    {
+      "epoch": 0.32306915699141847,
+      "grad_norm": 1.7867386563705459,
+      "learning_rate": 2.318939319175771e-05,
+      "loss": 0.6806,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3247518088507488,
+      "grad_norm": 2.519605910503542,
+      "learning_rate": 2.3122166529491822e-05,
+      "loss": 0.6837,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3264344607100791,
+      "grad_norm": 1.5090617010699425,
+      "learning_rate": 2.3054708248906483e-05,
+      "loss": 0.7201,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3281171125694094,
+      "grad_norm": 1.85373627743108,
+      "learning_rate": 2.2987020273696996e-05,
+      "loss": 0.7007,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3297997644287397,
+      "grad_norm": 3.1668783585579714,
+      "learning_rate": 2.2919104534108825e-05,
+      "loss": 0.6827,
+      "step": 1960
+    },
+    {
+      "epoch": 0.33148241628807,
+      "grad_norm": 2.802801151344103,
+      "learning_rate": 2.2850962966882547e-05,
+      "loss": 0.733,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3331650681474003,
+      "grad_norm": 4.351080547606847,
+      "learning_rate": 2.278259751519861e-05,
+      "loss": 0.7125,
+      "step": 1980
+    },
+    {
+      "epoch": 0.33484772000673063,
+      "grad_norm": 1.4284076903376268,
+      "learning_rate": 2.2714010128621957e-05,
+      "loss": 0.7166,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3365303718660609,
+      "grad_norm": 1.4047557097137526,
+      "learning_rate": 2.2645202763046385e-05,
+      "loss": 0.7306,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 5943,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 400,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.11190196748288e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-2000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e95a8f5e7f8a0f6f3e1f415e9606de2bf6f80315b55f9012ea921093e8d88264
+size 6520

checkpoint-2000/zero_to_fp32.py

@@ -0,0 +1,587 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example: python zero_to_fp32.py . pytorch_model.bin
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+
+    total_files = len(files)
+    state_dicts = []
+    for f in files:
+        state_dict = torch.load(f, map_location=device)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
+        ]
+
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel()
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in param_shapes.items():
+
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+    print(f"Saving fp32 state dict to {output_file}")
+    torch.save(state_dict, output_file)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument(
+        "output_file",
+        type=str,
+        help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, args.output_file, tag=args.tag)

checkpoint-2400/README.md

@@ -0,0 +1,202 @@
+---
+library_name: peft
+base_model: Qwen/Qwen-VL-Chat
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.10.0

checkpoint-2400/adapter_config.json

@@ -0,0 +1,380 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen-VL-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "transformer.h.17.attn.c_proj",
+    "transformer.h.20.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.1.attn.in_proj",
+    "transformer.h.3.attn.c_attn",
+    "transformer.visual.transformer.resblocks.12.attn.in_proj",
+    "transformer.visual.transformer.resblocks.47.attn.in_proj",
+    "transformer.h.28.mlp.w2",
+    "transformer.h.6.mlp.w2",
+    "transformer.h.13.mlp.w1",
+    "transformer.visual.transformer.resblocks.39.attn.out_proj",
+    "transformer.h.2.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.attn.out_proj",
+    "transformer.visual.transformer.resblocks.0.attn.out_proj",
+    "transformer.h.4.attn.c_proj",
+    "transformer.h.22.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.12.attn.out_proj",
+    "transformer.visual.transformer.resblocks.10.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.43.attn.in_proj",
+    "transformer.visual.transformer.resblocks.0.attn.in_proj",
+    "transformer.visual.transformer.resblocks.26.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.11.mlp.c_proj",
+    "transformer.h.0.attn.c_attn",
+    "transformer.h.19.mlp.w2",
+    "transformer.visual.transformer.resblocks.37.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.40.mlp.c_proj",
+    "transformer.h.31.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.mlp.c_fc",
+    "transformer.h.18.mlp.w1",
+    "transformer.h.23.mlp.w2",
+    "transformer.visual.transformer.resblocks.6.attn.out_proj",
+    "transformer.visual.transformer.resblocks.17.attn.in_proj",
+    "transformer.visual.transformer.resblocks.27.attn.out_proj",
+    "transformer.h.12.mlp.w2",
+    "transformer.h.23.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.29.attn.in_proj",
+    "transformer.h.10.mlp.w1",
+    "transformer.visual.transformer.resblocks.18.attn.out_proj",
+    "transformer.visual.transformer.resblocks.4.attn.out_proj",
+    "transformer.visual.transformer.resblocks.8.mlp.c_fc",
+    "transformer.h.9.mlp.w1",
+    "transformer.visual.transformer.resblocks.38.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.attn.in_proj",
+    "transformer.visual.transformer.resblocks.14.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.22.attn.in_proj",
+    "transformer.visual.transformer.resblocks.25.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.23.attn.out_proj",
+    "transformer.visual.transformer.resblocks.3.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.41.mlp.c_proj",
+    "transformer.h.24.attn.c_proj",
+    "transformer.visual.transformer.resblocks.7.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.38.mlp.c_fc",
+    "transformer.h.10.attn.c_attn",
+    "transformer.h.26.attn.c_attn",
+    "transformer.visual.transformer.resblocks.5.attn.in_proj",
+    "transformer.visual.transformer.resblocks.2.attn.out_proj",
+    "transformer.h.7.attn.c_proj",
+    "transformer.h.24.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.34.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.2.mlp.c_proj",
+    "transformer.h.12.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.14.attn.out_proj",
+    "transformer.h.18.attn.c_attn",
+    "transformer.h.23.attn.c_proj",
+    "transformer.h.27.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.26.mlp.c_proj",
+    "transformer.h.3.mlp.w1",
+    "transformer.h.2.mlp.w2",
+    "transformer.visual.transformer.resblocks.45.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.25.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.45.attn.out_proj",
+    "transformer.h.25.mlp.w1",
+    "transformer.visual.transformer.resblocks.15.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.24.attn.in_proj",
+    "transformer.h.1.attn.c_proj",
+    "transformer.h.1.attn.c_attn",
+    "transformer.visual.transformer.resblocks.4.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.9.mlp.c_fc",
+    "transformer.h.13.attn.c_attn",
+    "transformer.visual.transformer.resblocks.40.attn.out_proj",
+    "transformer.h.7.mlp.w2",
+    "transformer.h.9.attn.c_proj",
+    "transformer.h.15.attn.c_attn",
+    "transformer.visual.transformer.resblocks.0.mlp.c_fc",
+    "transformer.h.27.attn.c_attn",
+    "transformer.h.15.mlp.c_proj",
+    "transformer.h.21.mlp.w2",
+    "transformer.h.28.attn.c_proj",
+    "transformer.visual.transformer.resblocks.42.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.16.attn.out_proj",
+    "transformer.h.9.mlp.w2",
+    "transformer.visual.transformer.resblocks.9.attn.in_proj",
+    "transformer.visual.transformer.resblocks.28.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.mlp.c_proj",
+    "transformer.h.11.mlp.w1",
+    "transformer.visual.transformer.resblocks.18.attn.in_proj",
+    "transformer.h.10.attn.c_proj",
+    "transformer.visual.transformer.resblocks.42.mlp.c_fc",
+    "transformer.h.31.attn.c_attn",
+    "transformer.visual.transformer.resblocks.23.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.21.attn.in_proj",
+    "transformer.h.24.mlp.w1",
+    "transformer.visual.transformer.resblocks.35.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.7.mlp.c_proj",
+    "transformer.h.8.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.12.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.7.attn.out_proj",
+    "transformer.h.22.mlp.w2",
+    "transformer.h.29.mlp.w2",
+    "transformer.h.0.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.38.attn.in_proj",
+    "transformer.h.8.mlp.w1",
+    "transformer.h.0.mlp.w1",
+    "transformer.h.26.mlp.w2",
+    "transformer.h.25.attn.c_proj",
+    "transformer.h.27.mlp.w1",
+    "transformer.visual.transformer.resblocks.21.attn.out_proj",
+    "transformer.visual.transformer.resblocks.44.attn.in_proj",
+    "transformer.visual.transformer.resblocks.43.attn.out_proj",
+    "transformer.h.29.attn.c_attn",
+    "transformer.h.24.attn.c_attn",
+    "transformer.visual.transformer.resblocks.17.attn.out_proj",
+    "transformer.h.2.attn.c_proj",
+    "transformer.visual.transformer.resblocks.15.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.11.attn.in_proj",
+    "transformer.visual.transformer.resblocks.17.mlp.c_proj",
+    "transformer.h.11.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.6.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.41.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.5.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.18.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.24.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.attn.out_proj",
+    "transformer.h.1.mlp.w2",
+    "transformer.h.21.mlp.c_proj",
+    "transformer.h.23.attn.c_attn",
+    "transformer.visual.transformer.resblocks.34.attn.out_proj",
+    "transformer.h.14.attn.c_attn",
+    "transformer.visual.transformer.resblocks.2.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.31.attn.out_proj",
+    "transformer.visual.transformer.resblocks.30.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.11.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.31.attn.in_proj",
+    "transformer.visual.transformer.resblocks.39.mlp.c_proj",
+    "transformer.h.9.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.20.attn.out_proj",
+    "transformer.h.18.mlp.c_proj",
+    "transformer.h.19.mlp.w1",
+    "transformer.h.9.attn.c_attn",
+    "transformer.visual.transformer.resblocks.36.attn.out_proj",
+    "transformer.visual.transformer.resblocks.7.attn.in_proj",
+    "transformer.visual.transformer.resblocks.30.attn.in_proj",
+    "transformer.visual.transformer.resblocks.47.attn.out_proj",
+    "transformer.visual.transformer.resblocks.0.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.15.attn.in_proj",
+    "transformer.visual.transformer.resblocks.29.attn.out_proj",
+    "transformer.visual.transformer.resblocks.41.attn.in_proj",
+    "transformer.visual.transformer.resblocks.4.attn.in_proj",
+    "transformer.h.25.attn.c_attn",
+    "transformer.visual.transformer.resblocks.12.mlp.c_proj",
+    "transformer.h.16.mlp.w1",
+    "transformer.h.28.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.27.attn.in_proj",
+    "transformer.visual.transformer.resblocks.13.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.attn.in_proj",
+    "transformer.visual.transformer.resblocks.45.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.46.mlp.c_proj",
+    "transformer.h.30.mlp.w1",
+    "transformer.visual.transformer.resblocks.43.mlp.c_fc",
+    "transformer.h.15.mlp.w1",
+    "transformer.h.16.attn.c_proj",
+    "transformer.h.20.mlp.w1",
+    "transformer.visual.transformer.resblocks.21.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.10.mlp.c_proj",
+    "transformer.h.10.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.35.attn.in_proj",
+    "transformer.h.13.mlp.w2",
+    "transformer.visual.transformer.resblocks.8.attn.out_proj",
+    "transformer.visual.transformer.resblocks.20.mlp.c_proj",
+    "transformer.h.22.attn.c_proj",
+    "transformer.h.6.mlp.w1",
+    "transformer.h.18.mlp.w2",
+    "transformer.h.4.mlp.c_proj",
+    "transformer.h.3.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.42.attn.out_proj",
+    "transformer.visual.transformer.resblocks.36.attn.in_proj",
+    "transformer.visual.transformer.resblocks.17.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.43.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.37.attn.in_proj",
+    "transformer.visual.transformer.resblocks.1.attn.out_proj",
+    "transformer.visual.transformer.resblocks.22.mlp.c_fc",
+    "transformer.h.22.mlp.w1",
+    "transformer.visual.transformer.resblocks.44.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.37.attn.out_proj",
+    "transformer.visual.transformer.resblocks.34.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.29.mlp.c_fc",
+    "transformer.h.18.attn.c_proj",
+    "transformer.visual.transformer.resblocks.38.attn.out_proj",
+    "transformer.h.5.attn.c_attn",
+    "transformer.visual.transformer.resblocks.19.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.15.attn.out_proj",
+    "transformer.visual.transformer.resblocks.37.mlp.c_fc",
+    "transformer.h.5.attn.c_proj",
+    "transformer.h.7.attn.c_attn",
+    "transformer.visual.transformer.resblocks.28.attn.out_proj",
+    "transformer.visual.transformer.resblocks.31.mlp.c_proj",
+    "transformer.h.29.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.45.attn.in_proj",
+    "transformer.visual.transformer.resblocks.27.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.10.attn.out_proj",
+    "transformer.visual.transformer.resblocks.40.attn.in_proj",
+    "transformer.h.23.mlp.w1",
+    "transformer.visual.transformer.resblocks.28.attn.in_proj",
+    "transformer.h.12.attn.c_proj",
+    "transformer.h.16.mlp.w2",
+    "transformer.h.27.mlp.w2",
+    "transformer.visual.transformer.resblocks.22.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.9.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.47.mlp.c_proj",
+    "transformer.h.26.attn.c_proj",
+    "transformer.visual.transformer.resblocks.40.mlp.c_fc",
+    "transformer.h.8.mlp.w2",
+    "transformer.visual.transformer.resblocks.27.mlp.c_fc",
+    "transformer.h.17.mlp.w1",
+    "transformer.h.31.mlp.w2",
+    "transformer.visual.transformer.resblocks.11.attn.out_proj",
+    "transformer.h.28.mlp.w1",
+    "transformer.visual.transformer.resblocks.10.attn.in_proj",
+    "transformer.h.12.mlp.w1",
+    "transformer.h.30.mlp.w2",
+    "transformer.visual.transformer.resblocks.13.attn.in_proj",
+    "transformer.h.6.attn.c_attn",
+    "transformer.h.5.mlp.c_proj",
+    "transformer.h.6.mlp.c_proj",
+    "transformer.h.22.attn.c_attn",
+    "transformer.h.13.attn.c_proj",
+    "transformer.visual.transformer.resblocks.46.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.41.attn.out_proj",
+    "transformer.visual.transformer.resblocks.30.mlp.c_fc",
+    "transformer.h.17.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.5.attn.out_proj",
+    "transformer.h.4.mlp.w2",
+    "transformer.visual.transformer.resblocks.1.mlp.c_proj",
+    "transformer.h.11.mlp.w2",
+    "transformer.h.19.attn.c_attn",
+    "transformer.h.14.mlp.w1",
+    "transformer.visual.transformer.resblocks.44.attn.out_proj",
+    "transformer.visual.transformer.resblocks.14.mlp.c_fc",
+    "transformer.h.21.attn.c_attn",
+    "transformer.visual.transformer.resblocks.36.mlp.c_proj",
+    "transformer.h.2.mlp.w1",
+    "transformer.h.14.attn.c_proj",
+    "transformer.visual.transformer.resblocks.46.attn.in_proj",
+    "transformer.h.6.attn.c_proj",
+    "transformer.h.0.mlp.w2",
+    "transformer.h.5.mlp.w1",
+    "transformer.h.30.attn.c_proj",
+    "transformer.h.24.mlp.w2",
+    "transformer.h.0.attn.c_proj",
+    "transformer.visual.transformer.resblocks.4.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.22.attn.out_proj",
+    "transformer.h.10.mlp.w2",
+    "transformer.h.17.mlp.w2",
+    "transformer.visual.transformer.resblocks.23.attn.in_proj",
+    "transformer.visual.transformer.resblocks.36.mlp.c_fc",
+    "transformer.h.20.mlp.w2",
+    "transformer.visual.transformer.resblocks.9.attn.out_proj",
+    "transformer.h.29.mlp.w1",
+    "transformer.visual.transformer.resblocks.20.attn.in_proj",
+    "transformer.visual.transformer.resblocks.20.mlp.c_fc",
+    "transformer.h.15.attn.c_proj",
+    "transformer.h.3.mlp.w2",
+    "transformer.h.30.attn.c_attn",
+    "transformer.visual.transformer.resblocks.47.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.16.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.39.mlp.c_fc",
+    "transformer.h.20.attn.c_attn",
+    "transformer.h.19.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.46.attn.out_proj",
+    "transformer.visual.transformer.resblocks.29.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.19.attn.out_proj",
+    "transformer.visual.transformer.resblocks.26.attn.in_proj",
+    "transformer.visual.transformer.resblocks.16.mlp.c_fc",
+    "transformer.h.11.attn.c_proj",
+    "transformer.h.12.attn.c_attn",
+    "transformer.visual.conv1",
+    "transformer.visual.transformer.resblocks.35.attn.out_proj",
+    "transformer.h.25.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.14.attn.in_proj",
+    "transformer.h.26.mlp.w1",
+    "transformer.visual.transformer.resblocks.1.mlp.c_fc",
+    "transformer.h.7.mlp.c_proj",
+    "transformer.h.29.attn.c_proj",
+    "transformer.h.1.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.33.mlp.c_proj",
+    "transformer.h.14.mlp.c_proj",
+    "transformer.h.3.attn.c_proj",
+    "transformer.h.25.mlp.w2",
+    "transformer.h.20.attn.c_proj",
+    "transformer.h.16.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.attn.in_proj",
+    "transformer.h.17.attn.c_attn",
+    "transformer.h.14.mlp.w2",
+    "transformer.visual.transformer.resblocks.2.attn.in_proj",
+    "transformer.visual.transformer.resblocks.5.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.3.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.33.attn.out_proj",
+    "transformer.h.15.mlp.w2",
+    "transformer.h.4.attn.c_attn",
+    "transformer.h.31.mlp.w1",
+    "transformer.h.11.attn.c_attn",
+    "transformer.visual.transformer.resblocks.23.mlp.c_proj",
+    "transformer.h.7.mlp.w1",
+    "transformer.visual.transformer.resblocks.34.attn.in_proj",
+    "transformer.h.1.mlp.w1",
+    "transformer.visual.transformer.resblocks.28.mlp.c_fc",
+    "transformer.h.21.attn.c_proj",
+    "transformer.h.30.mlp.c_proj",
+    "transformer.h.21.mlp.w1",
+    "transformer.visual.transformer.resblocks.30.attn.out_proj",
+    "transformer.visual.transformer.resblocks.42.attn.in_proj",
+    "transformer.visual.transformer.resblocks.25.attn.out_proj",
+    "transformer.visual.transformer.resblocks.19.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.39.attn.in_proj",
+    "transformer.visual.transformer.resblocks.19.attn.in_proj",
+    "transformer.visual.transformer.resblocks.13.mlp.c_fc",
+    "transformer.h.13.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.25.attn.in_proj",
+    "transformer.visual.transformer.resblocks.31.mlp.c_fc",
+    "transformer.visual.transformer.resblocks.24.attn.out_proj",
+    "transformer.visual.transformer.resblocks.24.mlp.c_fc",
+    "transformer.h.4.mlp.w1",
+    "transformer.h.8.attn.c_attn",
+    "transformer.visual.transformer.resblocks.21.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.44.mlp.c_proj",
+    "transformer.h.28.attn.c_attn",
+    "transformer.visual.transformer.resblocks.18.mlp.c_proj",
+    "transformer.visual.transformer.resblocks.32.attn.in_proj",
+    "transformer.h.19.attn.c_proj",
+    "transformer.h.2.attn.c_attn",
+    "transformer.visual.transformer.resblocks.35.mlp.c_proj",
+    "transformer.h.26.mlp.c_proj",
+    "transformer.h.8.attn.c_proj",
+    "transformer.h.27.attn.c_proj",
+    "transformer.visual.transformer.resblocks.13.attn.out_proj",
+    "transformer.h.16.attn.c_attn",
+    "transformer.visual.transformer.resblocks.16.attn.in_proj",
+    "transformer.visual.transformer.resblocks.8.attn.in_proj",
+    "transformer.visual.transformer.resblocks.26.attn.out_proj",
+    "transformer.h.31.attn.c_proj",
+    "transformer.h.5.mlp.w2",
+    "transformer.visual.transformer.resblocks.8.mlp.c_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}