Upload folder using huggingface_hub

checkpoint-120/adapter_model/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0

checkpoint-120/adapter_model/lora_default/adapter_config.json

@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "/nobackup/users/zfchen/cdl/LLaVA-RLHF/VILA1.5-3b/llm",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16,
+  "lora_dropout": 0.0,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "k_proj",
+    "v_proj",
+    "o_proj",
+    "gate_proj",
+    "up_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

checkpoint-120/adapter_model/lora_default/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:091c4020c80e9d9fc315c44e8bf529be50c3430db06b6f8c68428f5a7c3ada74
+size 224393877

checkpoint-120/config.json

@@ -0,0 +1,10 @@
+{
+  "architectures": [
+    "RewardModel"
+  ],
+  "backbone_model_name_or_path": "/nobackup/users/zfchen/cdl/LLaVA-RLHF/VILA1.5-3b/llm",
+  "mm_use_im_start_end": false,
+  "model_type": "reward_model",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.30.2"
+}

checkpoint-120/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9108380eb8e76c1064593403bcea3798532ffc999f60aa83d77f1986c3111852
+size 21687

checkpoint-120/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23ccbfcfaa29f8718318d2ca568c15317eeb460c340c72d82fecc183d56d17ba
+size 21687

checkpoint-120/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02dd6b8354207613ffddba9ad82b40cd300d39a17a5b5bc7461d9a0e9dbbfb1e
+size 21687

checkpoint-120/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28db48b3220dbbb391220d436cc3aa0a9dc31b7f44552c0ab3cb87b13905d62a
+size 21687

checkpoint-120/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b76df961261b4c1808f6de8fd12ddfb9adaa71836e4192a0e5e0588b4d5011f
+size 21687

checkpoint-120/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fcaabcf235f388013089a0731f3185918b9bcf3efc3be8cafd3048d57e5229e2
+size 21687

checkpoint-120/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc00b4d49ee413bc73f995a9ce3c0ecd7fd6c1628c392094668f46d0635029d8
+size 21687

checkpoint-120/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8b14a8a29425691a14fc71d787608a6554eded1f16c908c18ada686cc97ec42
+size 21687

checkpoint-120/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f735eb293f032dfeb9e458f35501d617ae5cf159a29a9b6ac465d72d0667b44e
+size 627

checkpoint-120/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<unk>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-120/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7aedb3582ecda9fa99ee9242c17a9658f6744db083ee6ebdc8fb14857f84d220
+size 499723

checkpoint-120/tokenizer_config.json

@@ -0,0 +1,72 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "add_prefix_space": true,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": {
+    "__type": "AddedToken",
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "clean_up_tokenization_spaces": false,
+  "eos_token": {
+    "__type": "AddedToken",
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "legacy": false,
+  "model_max_length": 4096,
+  "pad_token": {
+    "__type": "AddedToken",
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "padding_side": "left",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "truncation_side": "right",
+  "unk_token": {
+    "__type": "AddedToken",
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "use_default_system_prompt": false
+}

checkpoint-120/trainer_state.json

@@ -0,0 +1,442 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.304932735426009,
+  "global_step": 120,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.7037037037037036e-07,
+      "loss": 0.6848,
+      "step": 2
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 0.6805,
+      "step": 4
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.6842,
+      "step": 6
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.6893,
+      "step": 8
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.6787,
+      "step": 10
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.6789,
+      "step": 12
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 0.685,
+      "step": 14
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 0.6823,
+      "step": 16
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.6836,
+      "step": 18
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.6756,
+      "step": 20
+    },
+    {
+      "epoch": 0.72,
+      "eval_accuracy": 0.5419999957084656,
+      "eval_average_score": 0.5289427638053894,
+      "eval_label_positive_rate": 0.47600001096725464,
+      "eval_loss": 0.6793817281723022,
+      "eval_runtime": 8.697,
+      "eval_samples_per_second": 57.491,
+      "eval_steps_per_second": 7.244,
+      "step": 20
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.074074074074074e-06,
+      "loss": 0.6845,
+      "step": 22
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.6837,
+      "step": 24
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 0.6743,
+      "step": 26
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.6718,
+      "step": 28
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.6777,
+      "step": 30
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.6702,
+      "step": 32
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 0.6743,
+      "step": 34
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.6722,
+      "step": 36
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.6697,
+      "step": 38
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.6654,
+      "step": 40
+    },
+    {
+      "epoch": 1.43,
+      "eval_accuracy": 0.7080000042915344,
+      "eval_average_score": 0.4319504201412201,
+      "eval_label_positive_rate": 0.47600001096725464,
+      "eval_loss": 0.6608079075813293,
+      "eval_runtime": 8.6968,
+      "eval_samples_per_second": 57.492,
+      "eval_steps_per_second": 7.244,
+      "step": 40
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 0.6572,
+      "step": 42
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.6621,
+      "step": 44
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 0.6575,
+      "step": 46
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.6455,
+      "step": 48
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 9.25925925925926e-06,
+      "loss": 0.6367,
+      "step": 50
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.64,
+      "step": 52
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1e-05,
+      "loss": 0.6316,
+      "step": 54
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1e-05,
+      "loss": 0.6121,
+      "step": 56
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1e-05,
+      "loss": 0.6104,
+      "step": 58
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1e-05,
+      "loss": 0.5916,
+      "step": 60
+    },
+    {
+      "epoch": 2.15,
+      "eval_accuracy": 0.9419999718666077,
+      "eval_average_score": 0.13356663286685944,
+      "eval_label_positive_rate": 0.47600001096725464,
+      "eval_loss": 0.5838245153427124,
+      "eval_runtime": 8.6962,
+      "eval_samples_per_second": 57.496,
+      "eval_steps_per_second": 7.245,
+      "step": 60
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1e-05,
+      "loss": 0.5801,
+      "step": 62
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1e-05,
+      "loss": 0.5677,
+      "step": 64
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1e-05,
+      "loss": 0.5472,
+      "step": 66
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1e-05,
+      "loss": 0.5267,
+      "step": 68
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1e-05,
+      "loss": 0.5224,
+      "step": 70
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1e-05,
+      "loss": 0.49,
+      "step": 72
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1e-05,
+      "loss": 0.4607,
+      "step": 74
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1e-05,
+      "loss": 0.4295,
+      "step": 76
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1e-05,
+      "loss": 0.393,
+      "step": 78
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3463,
+      "step": 80
+    },
+    {
+      "epoch": 2.87,
+      "eval_accuracy": 0.9800000190734863,
+      "eval_average_score": -0.13676369190216064,
+      "eval_label_positive_rate": 0.47600001096725464,
+      "eval_loss": 0.30906087160110474,
+      "eval_runtime": 8.6996,
+      "eval_samples_per_second": 57.474,
+      "eval_steps_per_second": 7.242,
+      "step": 80
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3167,
+      "step": 82
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1e-05,
+      "loss": 0.2559,
+      "step": 84
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1e-05,
+      "loss": 0.2119,
+      "step": 86
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1e-05,
+      "loss": 0.1874,
+      "step": 88
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1e-05,
+      "loss": 0.1407,
+      "step": 90
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1e-05,
+      "loss": 0.1037,
+      "step": 92
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1e-05,
+      "loss": 0.0717,
+      "step": 94
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0692,
+      "step": 96
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1e-05,
+      "loss": 0.0708,
+      "step": 98
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 1e-05,
+      "loss": 0.0426,
+      "step": 100
+    },
+    {
+      "epoch": 3.59,
+      "eval_accuracy": 0.9879999756813049,
+      "eval_average_score": -0.7325789332389832,
+      "eval_label_positive_rate": 0.47600001096725464,
+      "eval_loss": 0.03151561692357063,
+      "eval_runtime": 8.708,
+      "eval_samples_per_second": 57.418,
+      "eval_steps_per_second": 7.235,
+      "step": 100
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0315,
+      "step": 102
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1e-05,
+      "loss": 0.0227,
+      "step": 104
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1e-05,
+      "loss": 0.0272,
+      "step": 106
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1e-05,
+      "loss": 0.0203,
+      "step": 108
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1e-05,
+      "loss": 0.0151,
+      "step": 110
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1e-05,
+      "loss": 0.0199,
+      "step": 112
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 1e-05,
+      "loss": 0.0171,
+      "step": 114
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1e-05,
+      "loss": 0.0094,
+      "step": 116
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 1e-05,
+      "loss": 0.0166,
+      "step": 118
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0079,
+      "step": 120
+    },
+    {
+      "epoch": 4.3,
+      "eval_accuracy": 1.0,
+      "eval_average_score": -1.2080471515655518,
+      "eval_label_positive_rate": 0.47600001096725464,
+      "eval_loss": 0.009560000151395798,
+      "eval_runtime": 8.7049,
+      "eval_samples_per_second": 57.439,
+      "eval_steps_per_second": 7.237,
+      "step": 120
+    }
+  ],
+  "max_steps": 270,
+  "num_train_epochs": 10,
+  "total_flos": 3.914291024788193e+17,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-120/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57b146ab149226a1406ae4dbc206a2fc5340b734501c6a044f3f0daf846ec023
+size 4795