danielpark commited on
Commit
5610aae
1 Parent(s): 7f1a36e

init: weights

Browse files
README.md CHANGED
@@ -1,3 +1,171 @@
1
  ---
 
2
  license: apache-2.0
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ library_name: transformers
3
  license: apache-2.0
4
+ tags:
5
+ - jamba
6
+ - mamba
7
+ - moe
8
  ---
9
+
10
+ # Jamba-v0.1-9B
11
+
12
+ A dense version of [Jamba-v0.1](https://huggingface.co/ai21labs/Jamba-v0.1), which extracts the weights of the first expert.
13
+ It no longer uses MoE. Please refer to [this script](https://github.com/TechxGenus/Jamba-utils/blob/main/dense_downcycling.py) for details.
14
+ It can use single 3090/4090 for inference, and the usage method is exactly the same as Jamba-v0.1.
15
+
16
+ ---
17
+
18
+ # Original Model Card for Jamba
19
+
20
+ Jamba is a state-of-the-art, hybrid SSM-Transformer LLM. It delivers throughput gains over traditional Transformer-based models, while outperforming or matching the leading models of its size class on most common benchmarks.
21
+
22
+ Jamba is the first production-scale Mamba implementation, which opens up interesting research and application opportunities. While this initial experimentation shows encouraging gains, we expect these to be further enhanced with future optimizations and explorations.
23
+
24
+ This model card is for the base version of Jamba. It’s a pretrained, mixture-of-experts (MoE) generative text model, with 12B active parameters and a total of 52B parameters across all experts. It supports a 256K context length, and can fit up to 140K tokens on a single 80GB GPU.
25
+
26
+ For full details of this model please read the [release blog post](https://www.ai21.com/blog/announcing-jamba).
27
+
28
+ ## Model Details
29
+
30
+ - **Developed by:** [AI21](https://www.ai21.com)
31
+ - **Model type:** Joint Attention and Mamba (Jamba)
32
+ - **License:** Apache 2.0
33
+ - **Context length:** 256K
34
+ - **Knowledge cutoff date:** March 5, 2024
35
+
36
+ ## Usage
37
+ ### Presequities
38
+ Jamba requires you use `transformers` version 4.39.0 or higher:
39
+ ```bash
40
+ pip install transformers>=4.39.0
41
+ ```
42
+
43
+ In order to run optimized Mamba implementations, you first need to install `mamba-ssm` and `causal-conv1d`:
44
+ ```bash
45
+ pip install mamba-ssm causal-conv1d>=1.2.0
46
+ ```
47
+ You also have to have the model on a CUDA device.
48
+
49
+ You can run the model not using the optimized Mamba kernels, but it is **not** recommended as it will result in significantly lower latencies. In order to do that, you'll need to specify `use_mamba_kernels=False` when loading the model.
50
+
51
+ ### Run the model
52
+ Please note that, at the moment, `trust_remote_code=True` is required for running the new Jamba architecture.
53
+ ```python
54
+ from transformers import AutoModelForCausalLM, AutoTokenizer
55
+
56
+ model = AutoModelForCausalLM.from_pretrained("ai21labs/Jamba-v0.1",
57
+ trust_remote_code=True)
58
+ tokenizer = AutoTokenizer.from_pretrained("ai21labs/Jamba-v0.1")
59
+
60
+ input_ids = tokenizer("In the recent Super Bowl LVIII,", return_tensors='pt').to(model.device)["input_ids"]
61
+
62
+ outputs = model.generate(input_ids, max_new_tokens=216)
63
+
64
+ print(tokenizer.batch_decode(outputs))
65
+ # ["<|startoftext|>In the recent Super Bowl LVIII, the Kansas City Chiefs emerged victorious, defeating the San Francisco 49ers in a thrilling overtime showdown. The game was a nail-biter, with both teams showcasing their skills and determination.\n\nThe Chiefs, led by their star quarterback Patrick Mahomes, displayed their offensive prowess, while the 49ers, led by their strong defense, put up a tough fight. The game went into overtime, with the Chiefs ultimately securing the win with a touchdown.\n\nThe victory marked the Chiefs' second Super Bowl win in four years, solidifying their status as one of the top teams in the NFL. The game was a testament to the skill and talent of both teams, and a thrilling end to the NFL season.\n\nThe Super Bowl is not just about the game itself, but also about the halftime show and the commercials. This year's halftime show featured a star-studded lineup, including Usher, Alicia Keys, and Lil Jon. The show was a spectacle of music and dance, with the performers delivering an energetic and entertaining performance.\n"]
66
+ ```
67
+
68
+ <details>
69
+ <summary><strong>Loading the model in half precision</strong></summary>
70
+
71
+ The published checkpoint is saved in BF16. In order to load it into RAM in BF16/FP16, you need to specify `torch_dtype`:
72
+
73
+ ```python
74
+ from transformers import AutoModelForCausalLM
75
+ import torch
76
+ model = AutoModelForCausalLM.from_pretrained("ai21labs/Jamba-v0.1",
77
+ trust_remote_code=True,
78
+ torch_dtype=torch.bfloat16) # you can also use torch_dtype=torch.float16
79
+ ```
80
+
81
+ When using half precision, you can enable the [FlashAttention2](https://github.com/Dao-AILab/flash-attention) implementation of the Attention blocks. In order to use it, you also need the model on a CUDA device. Since in this precision the model is to big to fit on a single 80GB GPU, you'll also need to parallelize it using [accelerate](https://huggingface.co/docs/accelerate/index):
82
+ ```python
83
+ from transformers import AutoModelForCausalLM
84
+ import torch
85
+ model = AutoModelForCausalLM.from_pretrained("ai21labs/Jamba-v0.1",
86
+ trust_remote_code=True,
87
+ torch_dtype=torch.bfloat16,
88
+ attn_implementation="flash_attention_2",
89
+ device_map="auto")
90
+ ```
91
+
92
+ </details>
93
+ <details><summary><strong>Load the model in 8-bit</strong></summary>
94
+
95
+ **Using 8-bit precision, it is possible to fit up to 140K sequence lengths on a single 80GB GPU.** You can easily quantize the model to 8-bit using [bitsandbytes](https://huggingface.co/docs/bitsandbytes/index). In order to not degrade model quality, we recommend to exclude the Mamba blocks from the quantization:
96
+
97
+ ```python
98
+ from transformers import AutoModelForCausalLM, BitsAndBytesConfig
99
+ quantization_config = BitsAndBytesConfig(load_in_8bit=True,
100
+ llm_int8_skip_modules=["mamba"])
101
+ model = AutoModelForCausalLM.from_pretrained("ai21labs/Jamba-v0.1",
102
+ trust_remote_code=True,
103
+ torch_dtype=torch.bfloat16,
104
+ attn_implementation="flash_attention_2",
105
+ quantization_config=quantization_config)
106
+ ```
107
+ </details>
108
+
109
+ ### Fine-tuning example
110
+ Jamba is a base model that can be fine-tuned for custom solutions (including for chat/instruct versions). You can fine-tune it using any technique of your choice. Here is an example of fine-tuning with the [PEFT](https://huggingface.co/docs/peft/index) library:
111
+
112
+ ```python
113
+ from datasets import load_dataset
114
+ from trl import SFTTrainer
115
+ from peft import LoraConfig
116
+ from transformers import AutoTokenizer, AutoModelForCausalLM, TrainingArguments
117
+
118
+ tokenizer = AutoTokenizer.from_pretrained("ai21labs/Jamba-v0.1")
119
+ model = AutoModelForCausalLM.from_pretrained("ai21labs/Jamba-v0.1", trust_remote_code=True, device_map='auto')
120
+
121
+ dataset = load_dataset("Abirate/english_quotes", split="train")
122
+ training_args = TrainingArguments(
123
+ output_dir="./results",
124
+ num_train_epochs=3,
125
+ per_device_train_batch_size=4,
126
+ logging_dir='./logs',
127
+ logging_steps=10,
128
+ learning_rate=2e-3
129
+ )
130
+ lora_config = LoraConfig(
131
+ r=8,
132
+ target_modules=["embed_tokens", "x_proj", "in_proj", "out_proj"],
133
+ task_type="CAUSAL_LM",
134
+ bias="none"
135
+ )
136
+ trainer = SFTTrainer(
137
+ model=model,
138
+ tokenizer=tokenizer,
139
+ args=training_args,
140
+ peft_config=lora_config,
141
+ train_dataset=dataset,
142
+ dataset_text_field="quote",
143
+ )
144
+
145
+ trainer.train()
146
+ ```
147
+
148
+ ## Results on common benchmarks
149
+ | Benchmark | Score |
150
+ |--------------|:-----:|
151
+ | HellaSwag | 87.1% |
152
+ | Arc Challenge | 64.4% |
153
+ | WinoGrande | 82.5% |
154
+ | PIQA | 83.2% |
155
+ | MMLU | 67.4% |
156
+ | BBH | 45.4% |
157
+ | TruthfulQA | 46.4% |
158
+ | GSM8K (CoT) | 59.9% |
159
+
160
+ It's crucial that the 'BOS' token is added to all prompts, which might not be enabled by default in all eval frameworks.
161
+
162
+
163
+ ## Notice
164
+ Jamba is a pretrained base model and did not undergo any alignment for instruct/chat interactions.
165
+
166
+ As a base model, Jamba is intended for use as a foundation layer for fine tuning, training, and developing custom solutions. Jamba does not have safety moderation mechanisms and guardrails should be added for responsible and safe use.
167
+
168
+ ## About AI21
169
+ AI21 builds reliable, practical, and scalable AI solutions for the enterprise.
170
+
171
+ Jamba is the first in AI21’s new family of models, and the Instruct version of Jamba is available in beta via the [AI21 platform](https://www.ai21.com/studio).
config.json ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "JambaForCausalLM"
4
+ ],
5
+ "attention_dropout": 0.0,
6
+ "attn_layer_offset": 4,
7
+ "attn_layer_period": 8,
8
+ "auto_map": {
9
+ "AutoConfig": "configuration_jamba.JambaConfig",
10
+ "AutoModel": "modeling_jamba.JambaModel",
11
+ "AutoModelForCausalLM": "modeling_jamba.JambaForCausalLM",
12
+ "AutoModelForSequenceClassification": "model.JambaForSequenceClassification"
13
+ },
14
+ "bos_token_id": 1,
15
+ "calc_logits_for_entire_prompt": false,
16
+ "eos_token_id": 2,
17
+ "expert_layer_offset": 1,
18
+ "expert_layer_period": 2,
19
+ "hidden_act": "silu",
20
+ "hidden_size": 4096,
21
+ "initializer_range": 0.02,
22
+ "intermediate_size": 14336,
23
+ "mamba_conv_bias": true,
24
+ "mamba_d_conv": 4,
25
+ "mamba_d_state": 16,
26
+ "mamba_dt_rank": 256,
27
+ "mamba_expand": 2,
28
+ "mamba_inner_layernorms": true,
29
+ "mamba_proj_bias": false,
30
+ "model_type": "jamba",
31
+ "n_ctx": 262144,
32
+ "num_attention_heads": 32,
33
+ "num_experts": 1,
34
+ "num_experts_per_tok": 1,
35
+ "num_hidden_layers": 32,
36
+ "num_key_value_heads": 8,
37
+ "output_router_logits": false,
38
+ "pad_token_id": 0,
39
+ "rms_norm_eps": 1e-06,
40
+ "router_aux_loss_coef": 0.001,
41
+ "sliding_window": null,
42
+ "tie_word_embeddings": false,
43
+ "torch_dtype": "bfloat16",
44
+ "transformers_version": "4.39.3",
45
+ "use_cache": true,
46
+ "use_mamba_kernels": true,
47
+ "vocab_size": 65536
48
+ }
configuration_jamba.py ADDED
Binary file (11.2 kB). View file
 
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 2,
5
+ "pad_token_id": 0,
6
+ "transformers_version": "4.39.3"
7
+ }
model-00001-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2c60b5778b6856888ca24447b3880c39a777e523e5fb94dd18ed33d9532d01f0
3
+ size 4916251136
model-00002-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58aa641dcb763378138fd1b4151c0726c8e06969417f1ddaccbcb04a6de69c79
3
+ size 4941976032
model-00003-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45ab7424c95e53b2a769629468ecfb953cdc2f65790d5fa2acea4f27d567b538
3
+ size 4941976080
model-00004-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9698548b8ec18b2258676c33df63f0c96986aa9a5c0846decc31c58e9d8a1e5c
3
+ size 3788575776
model.safetensors.index.json ADDED
@@ -0,0 +1,522 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 18588720896
4
+ },
5
+ "weight_map": {
6
+ "lm_head.weight": "model-00004-of-00004.safetensors",
7
+ "model.embed_tokens.weight": "model-00001-of-00004.safetensors",
8
+ "model.final_layernorm.weight": "model-00004-of-00004.safetensors",
9
+ "model.layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors",
10
+ "model.layers.0.mamba.A_log": "model-00001-of-00004.safetensors",
11
+ "model.layers.0.mamba.B_layernorm.weight": "model-00001-of-00004.safetensors",
12
+ "model.layers.0.mamba.C_layernorm.weight": "model-00001-of-00004.safetensors",
13
+ "model.layers.0.mamba.D": "model-00001-of-00004.safetensors",
14
+ "model.layers.0.mamba.conv1d.bias": "model-00001-of-00004.safetensors",
15
+ "model.layers.0.mamba.conv1d.weight": "model-00001-of-00004.safetensors",
16
+ "model.layers.0.mamba.dt_layernorm.weight": "model-00001-of-00004.safetensors",
17
+ "model.layers.0.mamba.dt_proj.bias": "model-00001-of-00004.safetensors",
18
+ "model.layers.0.mamba.dt_proj.weight": "model-00001-of-00004.safetensors",
19
+ "model.layers.0.mamba.in_proj.weight": "model-00001-of-00004.safetensors",
20
+ "model.layers.0.mamba.out_proj.weight": "model-00001-of-00004.safetensors",
21
+ "model.layers.0.mamba.x_proj.weight": "model-00001-of-00004.safetensors",
22
+ "model.layers.0.moe.experts.0.down_proj.weight": "model-00001-of-00004.safetensors",
23
+ "model.layers.0.moe.experts.0.gate_proj.weight": "model-00001-of-00004.safetensors",
24
+ "model.layers.0.moe.experts.0.up_proj.weight": "model-00001-of-00004.safetensors",
25
+ "model.layers.0.pre_moe_layernorm.weight": "model-00001-of-00004.safetensors",
26
+ "model.layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors",
27
+ "model.layers.1.mamba.A_log": "model-00001-of-00004.safetensors",
28
+ "model.layers.1.mamba.B_layernorm.weight": "model-00001-of-00004.safetensors",
29
+ "model.layers.1.mamba.C_layernorm.weight": "model-00001-of-00004.safetensors",
30
+ "model.layers.1.mamba.D": "model-00001-of-00004.safetensors",
31
+ "model.layers.1.mamba.conv1d.bias": "model-00001-of-00004.safetensors",
32
+ "model.layers.1.mamba.conv1d.weight": "model-00001-of-00004.safetensors",
33
+ "model.layers.1.mamba.dt_layernorm.weight": "model-00001-of-00004.safetensors",
34
+ "model.layers.1.mamba.dt_proj.bias": "model-00001-of-00004.safetensors",
35
+ "model.layers.1.mamba.dt_proj.weight": "model-00001-of-00004.safetensors",
36
+ "model.layers.1.mamba.in_proj.weight": "model-00001-of-00004.safetensors",
37
+ "model.layers.1.mamba.out_proj.weight": "model-00001-of-00004.safetensors",
38
+ "model.layers.1.mamba.x_proj.weight": "model-00001-of-00004.safetensors",
39
+ "model.layers.1.moe.experts.0.down_proj.weight": "model-00001-of-00004.safetensors",
40
+ "model.layers.1.moe.experts.0.gate_proj.weight": "model-00001-of-00004.safetensors",
41
+ "model.layers.1.moe.experts.0.up_proj.weight": "model-00001-of-00004.safetensors",
42
+ "model.layers.1.pre_moe_layernorm.weight": "model-00001-of-00004.safetensors",
43
+ "model.layers.10.input_layernorm.weight": "model-00002-of-00004.safetensors",
44
+ "model.layers.10.mamba.A_log": "model-00002-of-00004.safetensors",
45
+ "model.layers.10.mamba.B_layernorm.weight": "model-00002-of-00004.safetensors",
46
+ "model.layers.10.mamba.C_layernorm.weight": "model-00002-of-00004.safetensors",
47
+ "model.layers.10.mamba.D": "model-00002-of-00004.safetensors",
48
+ "model.layers.10.mamba.conv1d.bias": "model-00002-of-00004.safetensors",
49
+ "model.layers.10.mamba.conv1d.weight": "model-00002-of-00004.safetensors",
50
+ "model.layers.10.mamba.dt_layernorm.weight": "model-00002-of-00004.safetensors",
51
+ "model.layers.10.mamba.dt_proj.bias": "model-00002-of-00004.safetensors",
52
+ "model.layers.10.mamba.dt_proj.weight": "model-00002-of-00004.safetensors",
53
+ "model.layers.10.mamba.in_proj.weight": "model-00002-of-00004.safetensors",
54
+ "model.layers.10.mamba.out_proj.weight": "model-00002-of-00004.safetensors",
55
+ "model.layers.10.mamba.x_proj.weight": "model-00002-of-00004.safetensors",
56
+ "model.layers.10.moe.experts.0.down_proj.weight": "model-00002-of-00004.safetensors",
57
+ "model.layers.10.moe.experts.0.gate_proj.weight": "model-00002-of-00004.safetensors",
58
+ "model.layers.10.moe.experts.0.up_proj.weight": "model-00002-of-00004.safetensors",
59
+ "model.layers.10.pre_moe_layernorm.weight": "model-00002-of-00004.safetensors",
60
+ "model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors",
61
+ "model.layers.11.mamba.A_log": "model-00002-of-00004.safetensors",
62
+ "model.layers.11.mamba.B_layernorm.weight": "model-00002-of-00004.safetensors",
63
+ "model.layers.11.mamba.C_layernorm.weight": "model-00002-of-00004.safetensors",
64
+ "model.layers.11.mamba.D": "model-00002-of-00004.safetensors",
65
+ "model.layers.11.mamba.conv1d.bias": "model-00002-of-00004.safetensors",
66
+ "model.layers.11.mamba.conv1d.weight": "model-00002-of-00004.safetensors",
67
+ "model.layers.11.mamba.dt_layernorm.weight": "model-00002-of-00004.safetensors",
68
+ "model.layers.11.mamba.dt_proj.bias": "model-00002-of-00004.safetensors",
69
+ "model.layers.11.mamba.dt_proj.weight": "model-00002-of-00004.safetensors",
70
+ "model.layers.11.mamba.in_proj.weight": "model-00002-of-00004.safetensors",
71
+ "model.layers.11.mamba.out_proj.weight": "model-00002-of-00004.safetensors",
72
+ "model.layers.11.mamba.x_proj.weight": "model-00002-of-00004.safetensors",
73
+ "model.layers.11.moe.experts.0.down_proj.weight": "model-00002-of-00004.safetensors",
74
+ "model.layers.11.moe.experts.0.gate_proj.weight": "model-00002-of-00004.safetensors",
75
+ "model.layers.11.moe.experts.0.up_proj.weight": "model-00002-of-00004.safetensors",
76
+ "model.layers.11.pre_moe_layernorm.weight": "model-00002-of-00004.safetensors",
77
+ "model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors",
78
+ "model.layers.12.moe.experts.0.down_proj.weight": "model-00002-of-00004.safetensors",
79
+ "model.layers.12.moe.experts.0.gate_proj.weight": "model-00002-of-00004.safetensors",
80
+ "model.layers.12.moe.experts.0.up_proj.weight": "model-00002-of-00004.safetensors",
81
+ "model.layers.12.pre_moe_layernorm.weight": "model-00002-of-00004.safetensors",
82
+ "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
83
+ "model.layers.12.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
84
+ "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
85
+ "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
86
+ "model.layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors",
87
+ "model.layers.13.mamba.A_log": "model-00002-of-00004.safetensors",
88
+ "model.layers.13.mamba.B_layernorm.weight": "model-00002-of-00004.safetensors",
89
+ "model.layers.13.mamba.C_layernorm.weight": "model-00002-of-00004.safetensors",
90
+ "model.layers.13.mamba.D": "model-00002-of-00004.safetensors",
91
+ "model.layers.13.mamba.conv1d.bias": "model-00002-of-00004.safetensors",
92
+ "model.layers.13.mamba.conv1d.weight": "model-00002-of-00004.safetensors",
93
+ "model.layers.13.mamba.dt_layernorm.weight": "model-00002-of-00004.safetensors",
94
+ "model.layers.13.mamba.dt_proj.bias": "model-00002-of-00004.safetensors",
95
+ "model.layers.13.mamba.dt_proj.weight": "model-00002-of-00004.safetensors",
96
+ "model.layers.13.mamba.in_proj.weight": "model-00002-of-00004.safetensors",
97
+ "model.layers.13.mamba.out_proj.weight": "model-00002-of-00004.safetensors",
98
+ "model.layers.13.mamba.x_proj.weight": "model-00002-of-00004.safetensors",
99
+ "model.layers.13.moe.experts.0.down_proj.weight": "model-00002-of-00004.safetensors",
100
+ "model.layers.13.moe.experts.0.gate_proj.weight": "model-00002-of-00004.safetensors",
101
+ "model.layers.13.moe.experts.0.up_proj.weight": "model-00002-of-00004.safetensors",
102
+ "model.layers.13.pre_moe_layernorm.weight": "model-00002-of-00004.safetensors",
103
+ "model.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors",
104
+ "model.layers.14.mamba.A_log": "model-00002-of-00004.safetensors",
105
+ "model.layers.14.mamba.B_layernorm.weight": "model-00002-of-00004.safetensors",
106
+ "model.layers.14.mamba.C_layernorm.weight": "model-00002-of-00004.safetensors",
107
+ "model.layers.14.mamba.D": "model-00002-of-00004.safetensors",
108
+ "model.layers.14.mamba.conv1d.bias": "model-00002-of-00004.safetensors",
109
+ "model.layers.14.mamba.conv1d.weight": "model-00002-of-00004.safetensors",
110
+ "model.layers.14.mamba.dt_layernorm.weight": "model-00002-of-00004.safetensors",
111
+ "model.layers.14.mamba.dt_proj.bias": "model-00002-of-00004.safetensors",
112
+ "model.layers.14.mamba.dt_proj.weight": "model-00002-of-00004.safetensors",
113
+ "model.layers.14.mamba.in_proj.weight": "model-00002-of-00004.safetensors",
114
+ "model.layers.14.mamba.out_proj.weight": "model-00002-of-00004.safetensors",
115
+ "model.layers.14.mamba.x_proj.weight": "model-00002-of-00004.safetensors",
116
+ "model.layers.14.moe.experts.0.down_proj.weight": "model-00002-of-00004.safetensors",
117
+ "model.layers.14.moe.experts.0.gate_proj.weight": "model-00002-of-00004.safetensors",
118
+ "model.layers.14.moe.experts.0.up_proj.weight": "model-00002-of-00004.safetensors",
119
+ "model.layers.14.pre_moe_layernorm.weight": "model-00002-of-00004.safetensors",
120
+ "model.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors",
121
+ "model.layers.15.mamba.A_log": "model-00002-of-00004.safetensors",
122
+ "model.layers.15.mamba.B_layernorm.weight": "model-00002-of-00004.safetensors",
123
+ "model.layers.15.mamba.C_layernorm.weight": "model-00002-of-00004.safetensors",
124
+ "model.layers.15.mamba.D": "model-00002-of-00004.safetensors",
125
+ "model.layers.15.mamba.conv1d.bias": "model-00002-of-00004.safetensors",
126
+ "model.layers.15.mamba.conv1d.weight": "model-00002-of-00004.safetensors",
127
+ "model.layers.15.mamba.dt_layernorm.weight": "model-00002-of-00004.safetensors",
128
+ "model.layers.15.mamba.dt_proj.bias": "model-00002-of-00004.safetensors",
129
+ "model.layers.15.mamba.dt_proj.weight": "model-00002-of-00004.safetensors",
130
+ "model.layers.15.mamba.in_proj.weight": "model-00002-of-00004.safetensors",
131
+ "model.layers.15.mamba.out_proj.weight": "model-00002-of-00004.safetensors",
132
+ "model.layers.15.mamba.x_proj.weight": "model-00002-of-00004.safetensors",
133
+ "model.layers.15.moe.experts.0.down_proj.weight": "model-00002-of-00004.safetensors",
134
+ "model.layers.15.moe.experts.0.gate_proj.weight": "model-00002-of-00004.safetensors",
135
+ "model.layers.15.moe.experts.0.up_proj.weight": "model-00002-of-00004.safetensors",
136
+ "model.layers.15.pre_moe_layernorm.weight": "model-00002-of-00004.safetensors",
137
+ "model.layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors",
138
+ "model.layers.16.mamba.A_log": "model-00002-of-00004.safetensors",
139
+ "model.layers.16.mamba.B_layernorm.weight": "model-00002-of-00004.safetensors",
140
+ "model.layers.16.mamba.C_layernorm.weight": "model-00002-of-00004.safetensors",
141
+ "model.layers.16.mamba.D": "model-00002-of-00004.safetensors",
142
+ "model.layers.16.mamba.conv1d.bias": "model-00002-of-00004.safetensors",
143
+ "model.layers.16.mamba.conv1d.weight": "model-00002-of-00004.safetensors",
144
+ "model.layers.16.mamba.dt_layernorm.weight": "model-00002-of-00004.safetensors",
145
+ "model.layers.16.mamba.dt_proj.bias": "model-00002-of-00004.safetensors",
146
+ "model.layers.16.mamba.dt_proj.weight": "model-00002-of-00004.safetensors",
147
+ "model.layers.16.mamba.in_proj.weight": "model-00002-of-00004.safetensors",
148
+ "model.layers.16.mamba.out_proj.weight": "model-00002-of-00004.safetensors",
149
+ "model.layers.16.mamba.x_proj.weight": "model-00002-of-00004.safetensors",
150
+ "model.layers.16.moe.experts.0.down_proj.weight": "model-00002-of-00004.safetensors",
151
+ "model.layers.16.moe.experts.0.gate_proj.weight": "model-00002-of-00004.safetensors",
152
+ "model.layers.16.moe.experts.0.up_proj.weight": "model-00002-of-00004.safetensors",
153
+ "model.layers.16.pre_moe_layernorm.weight": "model-00002-of-00004.safetensors",
154
+ "model.layers.17.input_layernorm.weight": "model-00003-of-00004.safetensors",
155
+ "model.layers.17.mamba.A_log": "model-00002-of-00004.safetensors",
156
+ "model.layers.17.mamba.B_layernorm.weight": "model-00003-of-00004.safetensors",
157
+ "model.layers.17.mamba.C_layernorm.weight": "model-00003-of-00004.safetensors",
158
+ "model.layers.17.mamba.D": "model-00002-of-00004.safetensors",
159
+ "model.layers.17.mamba.conv1d.bias": "model-00002-of-00004.safetensors",
160
+ "model.layers.17.mamba.conv1d.weight": "model-00002-of-00004.safetensors",
161
+ "model.layers.17.mamba.dt_layernorm.weight": "model-00003-of-00004.safetensors",
162
+ "model.layers.17.mamba.dt_proj.bias": "model-00003-of-00004.safetensors",
163
+ "model.layers.17.mamba.dt_proj.weight": "model-00003-of-00004.safetensors",
164
+ "model.layers.17.mamba.in_proj.weight": "model-00003-of-00004.safetensors",
165
+ "model.layers.17.mamba.out_proj.weight": "model-00003-of-00004.safetensors",
166
+ "model.layers.17.mamba.x_proj.weight": "model-00003-of-00004.safetensors",
167
+ "model.layers.17.moe.experts.0.down_proj.weight": "model-00003-of-00004.safetensors",
168
+ "model.layers.17.moe.experts.0.gate_proj.weight": "model-00003-of-00004.safetensors",
169
+ "model.layers.17.moe.experts.0.up_proj.weight": "model-00003-of-00004.safetensors",
170
+ "model.layers.17.pre_moe_layernorm.weight": "model-00003-of-00004.safetensors",
171
+ "model.layers.18.input_layernorm.weight": "model-00003-of-00004.safetensors",
172
+ "model.layers.18.mamba.A_log": "model-00003-of-00004.safetensors",
173
+ "model.layers.18.mamba.B_layernorm.weight": "model-00003-of-00004.safetensors",
174
+ "model.layers.18.mamba.C_layernorm.weight": "model-00003-of-00004.safetensors",
175
+ "model.layers.18.mamba.D": "model-00003-of-00004.safetensors",
176
+ "model.layers.18.mamba.conv1d.bias": "model-00003-of-00004.safetensors",
177
+ "model.layers.18.mamba.conv1d.weight": "model-00003-of-00004.safetensors",
178
+ "model.layers.18.mamba.dt_layernorm.weight": "model-00003-of-00004.safetensors",
179
+ "model.layers.18.mamba.dt_proj.bias": "model-00003-of-00004.safetensors",
180
+ "model.layers.18.mamba.dt_proj.weight": "model-00003-of-00004.safetensors",
181
+ "model.layers.18.mamba.in_proj.weight": "model-00003-of-00004.safetensors",
182
+ "model.layers.18.mamba.out_proj.weight": "model-00003-of-00004.safetensors",
183
+ "model.layers.18.mamba.x_proj.weight": "model-00003-of-00004.safetensors",
184
+ "model.layers.18.moe.experts.0.down_proj.weight": "model-00003-of-00004.safetensors",
185
+ "model.layers.18.moe.experts.0.gate_proj.weight": "model-00003-of-00004.safetensors",
186
+ "model.layers.18.moe.experts.0.up_proj.weight": "model-00003-of-00004.safetensors",
187
+ "model.layers.18.pre_moe_layernorm.weight": "model-00003-of-00004.safetensors",
188
+ "model.layers.19.input_layernorm.weight": "model-00003-of-00004.safetensors",
189
+ "model.layers.19.mamba.A_log": "model-00003-of-00004.safetensors",
190
+ "model.layers.19.mamba.B_layernorm.weight": "model-00003-of-00004.safetensors",
191
+ "model.layers.19.mamba.C_layernorm.weight": "model-00003-of-00004.safetensors",
192
+ "model.layers.19.mamba.D": "model-00003-of-00004.safetensors",
193
+ "model.layers.19.mamba.conv1d.bias": "model-00003-of-00004.safetensors",
194
+ "model.layers.19.mamba.conv1d.weight": "model-00003-of-00004.safetensors",
195
+ "model.layers.19.mamba.dt_layernorm.weight": "model-00003-of-00004.safetensors",
196
+ "model.layers.19.mamba.dt_proj.bias": "model-00003-of-00004.safetensors",
197
+ "model.layers.19.mamba.dt_proj.weight": "model-00003-of-00004.safetensors",
198
+ "model.layers.19.mamba.in_proj.weight": "model-00003-of-00004.safetensors",
199
+ "model.layers.19.mamba.out_proj.weight": "model-00003-of-00004.safetensors",
200
+ "model.layers.19.mamba.x_proj.weight": "model-00003-of-00004.safetensors",
201
+ "model.layers.19.moe.experts.0.down_proj.weight": "model-00003-of-00004.safetensors",
202
+ "model.layers.19.moe.experts.0.gate_proj.weight": "model-00003-of-00004.safetensors",
203
+ "model.layers.19.moe.experts.0.up_proj.weight": "model-00003-of-00004.safetensors",
204
+ "model.layers.19.pre_moe_layernorm.weight": "model-00003-of-00004.safetensors",
205
+ "model.layers.2.input_layernorm.weight": "model-00001-of-00004.safetensors",
206
+ "model.layers.2.mamba.A_log": "model-00001-of-00004.safetensors",
207
+ "model.layers.2.mamba.B_layernorm.weight": "model-00001-of-00004.safetensors",
208
+ "model.layers.2.mamba.C_layernorm.weight": "model-00001-of-00004.safetensors",
209
+ "model.layers.2.mamba.D": "model-00001-of-00004.safetensors",
210
+ "model.layers.2.mamba.conv1d.bias": "model-00001-of-00004.safetensors",
211
+ "model.layers.2.mamba.conv1d.weight": "model-00001-of-00004.safetensors",
212
+ "model.layers.2.mamba.dt_layernorm.weight": "model-00001-of-00004.safetensors",
213
+ "model.layers.2.mamba.dt_proj.bias": "model-00001-of-00004.safetensors",
214
+ "model.layers.2.mamba.dt_proj.weight": "model-00001-of-00004.safetensors",
215
+ "model.layers.2.mamba.in_proj.weight": "model-00001-of-00004.safetensors",
216
+ "model.layers.2.mamba.out_proj.weight": "model-00001-of-00004.safetensors",
217
+ "model.layers.2.mamba.x_proj.weight": "model-00001-of-00004.safetensors",
218
+ "model.layers.2.moe.experts.0.down_proj.weight": "model-00001-of-00004.safetensors",
219
+ "model.layers.2.moe.experts.0.gate_proj.weight": "model-00001-of-00004.safetensors",
220
+ "model.layers.2.moe.experts.0.up_proj.weight": "model-00001-of-00004.safetensors",
221
+ "model.layers.2.pre_moe_layernorm.weight": "model-00001-of-00004.safetensors",
222
+ "model.layers.20.input_layernorm.weight": "model-00003-of-00004.safetensors",
223
+ "model.layers.20.moe.experts.0.down_proj.weight": "model-00003-of-00004.safetensors",
224
+ "model.layers.20.moe.experts.0.gate_proj.weight": "model-00003-of-00004.safetensors",
225
+ "model.layers.20.moe.experts.0.up_proj.weight": "model-00003-of-00004.safetensors",
226
+ "model.layers.20.pre_moe_layernorm.weight": "model-00003-of-00004.safetensors",
227
+ "model.layers.20.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
228
+ "model.layers.20.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
229
+ "model.layers.20.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
230
+ "model.layers.20.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
231
+ "model.layers.21.input_layernorm.weight": "model-00003-of-00004.safetensors",
232
+ "model.layers.21.mamba.A_log": "model-00003-of-00004.safetensors",
233
+ "model.layers.21.mamba.B_layernorm.weight": "model-00003-of-00004.safetensors",
234
+ "model.layers.21.mamba.C_layernorm.weight": "model-00003-of-00004.safetensors",
235
+ "model.layers.21.mamba.D": "model-00003-of-00004.safetensors",
236
+ "model.layers.21.mamba.conv1d.bias": "model-00003-of-00004.safetensors",
237
+ "model.layers.21.mamba.conv1d.weight": "model-00003-of-00004.safetensors",
238
+ "model.layers.21.mamba.dt_layernorm.weight": "model-00003-of-00004.safetensors",
239
+ "model.layers.21.mamba.dt_proj.bias": "model-00003-of-00004.safetensors",
240
+ "model.layers.21.mamba.dt_proj.weight": "model-00003-of-00004.safetensors",
241
+ "model.layers.21.mamba.in_proj.weight": "model-00003-of-00004.safetensors",
242
+ "model.layers.21.mamba.out_proj.weight": "model-00003-of-00004.safetensors",
243
+ "model.layers.21.mamba.x_proj.weight": "model-00003-of-00004.safetensors",
244
+ "model.layers.21.moe.experts.0.down_proj.weight": "model-00003-of-00004.safetensors",
245
+ "model.layers.21.moe.experts.0.gate_proj.weight": "model-00003-of-00004.safetensors",
246
+ "model.layers.21.moe.experts.0.up_proj.weight": "model-00003-of-00004.safetensors",
247
+ "model.layers.21.pre_moe_layernorm.weight": "model-00003-of-00004.safetensors",
248
+ "model.layers.22.input_layernorm.weight": "model-00003-of-00004.safetensors",
249
+ "model.layers.22.mamba.A_log": "model-00003-of-00004.safetensors",
250
+ "model.layers.22.mamba.B_layernorm.weight": "model-00003-of-00004.safetensors",
251
+ "model.layers.22.mamba.C_layernorm.weight": "model-00003-of-00004.safetensors",
252
+ "model.layers.22.mamba.D": "model-00003-of-00004.safetensors",
253
+ "model.layers.22.mamba.conv1d.bias": "model-00003-of-00004.safetensors",
254
+ "model.layers.22.mamba.conv1d.weight": "model-00003-of-00004.safetensors",
255
+ "model.layers.22.mamba.dt_layernorm.weight": "model-00003-of-00004.safetensors",
256
+ "model.layers.22.mamba.dt_proj.bias": "model-00003-of-00004.safetensors",
257
+ "model.layers.22.mamba.dt_proj.weight": "model-00003-of-00004.safetensors",
258
+ "model.layers.22.mamba.in_proj.weight": "model-00003-of-00004.safetensors",
259
+ "model.layers.22.mamba.out_proj.weight": "model-00003-of-00004.safetensors",
260
+ "model.layers.22.mamba.x_proj.weight": "model-00003-of-00004.safetensors",
261
+ "model.layers.22.moe.experts.0.down_proj.weight": "model-00003-of-00004.safetensors",
262
+ "model.layers.22.moe.experts.0.gate_proj.weight": "model-00003-of-00004.safetensors",
263
+ "model.layers.22.moe.experts.0.up_proj.weight": "model-00003-of-00004.safetensors",
264
+ "model.layers.22.pre_moe_layernorm.weight": "model-00003-of-00004.safetensors",
265
+ "model.layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors",
266
+ "model.layers.23.mamba.A_log": "model-00003-of-00004.safetensors",
267
+ "model.layers.23.mamba.B_layernorm.weight": "model-00003-of-00004.safetensors",
268
+ "model.layers.23.mamba.C_layernorm.weight": "model-00003-of-00004.safetensors",
269
+ "model.layers.23.mamba.D": "model-00003-of-00004.safetensors",
270
+ "model.layers.23.mamba.conv1d.bias": "model-00003-of-00004.safetensors",
271
+ "model.layers.23.mamba.conv1d.weight": "model-00003-of-00004.safetensors",
272
+ "model.layers.23.mamba.dt_layernorm.weight": "model-00003-of-00004.safetensors",
273
+ "model.layers.23.mamba.dt_proj.bias": "model-00003-of-00004.safetensors",
274
+ "model.layers.23.mamba.dt_proj.weight": "model-00003-of-00004.safetensors",
275
+ "model.layers.23.mamba.in_proj.weight": "model-00003-of-00004.safetensors",
276
+ "model.layers.23.mamba.out_proj.weight": "model-00003-of-00004.safetensors",
277
+ "model.layers.23.mamba.x_proj.weight": "model-00003-of-00004.safetensors",
278
+ "model.layers.23.moe.experts.0.down_proj.weight": "model-00003-of-00004.safetensors",
279
+ "model.layers.23.moe.experts.0.gate_proj.weight": "model-00003-of-00004.safetensors",
280
+ "model.layers.23.moe.experts.0.up_proj.weight": "model-00003-of-00004.safetensors",
281
+ "model.layers.23.pre_moe_layernorm.weight": "model-00003-of-00004.safetensors",
282
+ "model.layers.24.input_layernorm.weight": "model-00003-of-00004.safetensors",
283
+ "model.layers.24.mamba.A_log": "model-00003-of-00004.safetensors",
284
+ "model.layers.24.mamba.B_layernorm.weight": "model-00003-of-00004.safetensors",
285
+ "model.layers.24.mamba.C_layernorm.weight": "model-00003-of-00004.safetensors",
286
+ "model.layers.24.mamba.D": "model-00003-of-00004.safetensors",
287
+ "model.layers.24.mamba.conv1d.bias": "model-00003-of-00004.safetensors",
288
+ "model.layers.24.mamba.conv1d.weight": "model-00003-of-00004.safetensors",
289
+ "model.layers.24.mamba.dt_layernorm.weight": "model-00003-of-00004.safetensors",
290
+ "model.layers.24.mamba.dt_proj.bias": "model-00003-of-00004.safetensors",
291
+ "model.layers.24.mamba.dt_proj.weight": "model-00003-of-00004.safetensors",
292
+ "model.layers.24.mamba.in_proj.weight": "model-00003-of-00004.safetensors",
293
+ "model.layers.24.mamba.out_proj.weight": "model-00003-of-00004.safetensors",
294
+ "model.layers.24.mamba.x_proj.weight": "model-00003-of-00004.safetensors",
295
+ "model.layers.24.moe.experts.0.down_proj.weight": "model-00003-of-00004.safetensors",
296
+ "model.layers.24.moe.experts.0.gate_proj.weight": "model-00003-of-00004.safetensors",
297
+ "model.layers.24.moe.experts.0.up_proj.weight": "model-00003-of-00004.safetensors",
298
+ "model.layers.24.pre_moe_layernorm.weight": "model-00003-of-00004.safetensors",
299
+ "model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors",
300
+ "model.layers.25.mamba.A_log": "model-00003-of-00004.safetensors",
301
+ "model.layers.25.mamba.B_layernorm.weight": "model-00003-of-00004.safetensors",
302
+ "model.layers.25.mamba.C_layernorm.weight": "model-00003-of-00004.safetensors",
303
+ "model.layers.25.mamba.D": "model-00003-of-00004.safetensors",
304
+ "model.layers.25.mamba.conv1d.bias": "model-00003-of-00004.safetensors",
305
+ "model.layers.25.mamba.conv1d.weight": "model-00003-of-00004.safetensors",
306
+ "model.layers.25.mamba.dt_layernorm.weight": "model-00003-of-00004.safetensors",
307
+ "model.layers.25.mamba.dt_proj.bias": "model-00003-of-00004.safetensors",
308
+ "model.layers.25.mamba.dt_proj.weight": "model-00003-of-00004.safetensors",
309
+ "model.layers.25.mamba.in_proj.weight": "model-00003-of-00004.safetensors",
310
+ "model.layers.25.mamba.out_proj.weight": "model-00003-of-00004.safetensors",
311
+ "model.layers.25.mamba.x_proj.weight": "model-00003-of-00004.safetensors",
312
+ "model.layers.25.moe.experts.0.down_proj.weight": "model-00003-of-00004.safetensors",
313
+ "model.layers.25.moe.experts.0.gate_proj.weight": "model-00003-of-00004.safetensors",
314
+ "model.layers.25.moe.experts.0.up_proj.weight": "model-00003-of-00004.safetensors",
315
+ "model.layers.25.pre_moe_layernorm.weight": "model-00003-of-00004.safetensors",
316
+ "model.layers.26.input_layernorm.weight": "model-00004-of-00004.safetensors",
317
+ "model.layers.26.mamba.A_log": "model-00003-of-00004.safetensors",
318
+ "model.layers.26.mamba.B_layernorm.weight": "model-00004-of-00004.safetensors",
319
+ "model.layers.26.mamba.C_layernorm.weight": "model-00004-of-00004.safetensors",
320
+ "model.layers.26.mamba.D": "model-00003-of-00004.safetensors",
321
+ "model.layers.26.mamba.conv1d.bias": "model-00003-of-00004.safetensors",
322
+ "model.layers.26.mamba.conv1d.weight": "model-00003-of-00004.safetensors",
323
+ "model.layers.26.mamba.dt_layernorm.weight": "model-00004-of-00004.safetensors",
324
+ "model.layers.26.mamba.dt_proj.bias": "model-00004-of-00004.safetensors",
325
+ "model.layers.26.mamba.dt_proj.weight": "model-00004-of-00004.safetensors",
326
+ "model.layers.26.mamba.in_proj.weight": "model-00004-of-00004.safetensors",
327
+ "model.layers.26.mamba.out_proj.weight": "model-00004-of-00004.safetensors",
328
+ "model.layers.26.mamba.x_proj.weight": "model-00004-of-00004.safetensors",
329
+ "model.layers.26.moe.experts.0.down_proj.weight": "model-00004-of-00004.safetensors",
330
+ "model.layers.26.moe.experts.0.gate_proj.weight": "model-00004-of-00004.safetensors",
331
+ "model.layers.26.moe.experts.0.up_proj.weight": "model-00004-of-00004.safetensors",
332
+ "model.layers.26.pre_moe_layernorm.weight": "model-00004-of-00004.safetensors",
333
+ "model.layers.27.input_layernorm.weight": "model-00004-of-00004.safetensors",
334
+ "model.layers.27.mamba.A_log": "model-00004-of-00004.safetensors",
335
+ "model.layers.27.mamba.B_layernorm.weight": "model-00004-of-00004.safetensors",
336
+ "model.layers.27.mamba.C_layernorm.weight": "model-00004-of-00004.safetensors",
337
+ "model.layers.27.mamba.D": "model-00004-of-00004.safetensors",
338
+ "model.layers.27.mamba.conv1d.bias": "model-00004-of-00004.safetensors",
339
+ "model.layers.27.mamba.conv1d.weight": "model-00004-of-00004.safetensors",
340
+ "model.layers.27.mamba.dt_layernorm.weight": "model-00004-of-00004.safetensors",
341
+ "model.layers.27.mamba.dt_proj.bias": "model-00004-of-00004.safetensors",
342
+ "model.layers.27.mamba.dt_proj.weight": "model-00004-of-00004.safetensors",
343
+ "model.layers.27.mamba.in_proj.weight": "model-00004-of-00004.safetensors",
344
+ "model.layers.27.mamba.out_proj.weight": "model-00004-of-00004.safetensors",
345
+ "model.layers.27.mamba.x_proj.weight": "model-00004-of-00004.safetensors",
346
+ "model.layers.27.moe.experts.0.down_proj.weight": "model-00004-of-00004.safetensors",
347
+ "model.layers.27.moe.experts.0.gate_proj.weight": "model-00004-of-00004.safetensors",
348
+ "model.layers.27.moe.experts.0.up_proj.weight": "model-00004-of-00004.safetensors",
349
+ "model.layers.27.pre_moe_layernorm.weight": "model-00004-of-00004.safetensors",
350
+ "model.layers.28.input_layernorm.weight": "model-00004-of-00004.safetensors",
351
+ "model.layers.28.moe.experts.0.down_proj.weight": "model-00004-of-00004.safetensors",
352
+ "model.layers.28.moe.experts.0.gate_proj.weight": "model-00004-of-00004.safetensors",
353
+ "model.layers.28.moe.experts.0.up_proj.weight": "model-00004-of-00004.safetensors",
354
+ "model.layers.28.pre_moe_layernorm.weight": "model-00004-of-00004.safetensors",
355
+ "model.layers.28.self_attn.k_proj.weight": "model-00004-of-00004.safetensors",
356
+ "model.layers.28.self_attn.o_proj.weight": "model-00004-of-00004.safetensors",
357
+ "model.layers.28.self_attn.q_proj.weight": "model-00004-of-00004.safetensors",
358
+ "model.layers.28.self_attn.v_proj.weight": "model-00004-of-00004.safetensors",
359
+ "model.layers.29.input_layernorm.weight": "model-00004-of-00004.safetensors",
360
+ "model.layers.29.mamba.A_log": "model-00004-of-00004.safetensors",
361
+ "model.layers.29.mamba.B_layernorm.weight": "model-00004-of-00004.safetensors",
362
+ "model.layers.29.mamba.C_layernorm.weight": "model-00004-of-00004.safetensors",
363
+ "model.layers.29.mamba.D": "model-00004-of-00004.safetensors",
364
+ "model.layers.29.mamba.conv1d.bias": "model-00004-of-00004.safetensors",
365
+ "model.layers.29.mamba.conv1d.weight": "model-00004-of-00004.safetensors",
366
+ "model.layers.29.mamba.dt_layernorm.weight": "model-00004-of-00004.safetensors",
367
+ "model.layers.29.mamba.dt_proj.bias": "model-00004-of-00004.safetensors",
368
+ "model.layers.29.mamba.dt_proj.weight": "model-00004-of-00004.safetensors",
369
+ "model.layers.29.mamba.in_proj.weight": "model-00004-of-00004.safetensors",
370
+ "model.layers.29.mamba.out_proj.weight": "model-00004-of-00004.safetensors",
371
+ "model.layers.29.mamba.x_proj.weight": "model-00004-of-00004.safetensors",
372
+ "model.layers.29.moe.experts.0.down_proj.weight": "model-00004-of-00004.safetensors",
373
+ "model.layers.29.moe.experts.0.gate_proj.weight": "model-00004-of-00004.safetensors",
374
+ "model.layers.29.moe.experts.0.up_proj.weight": "model-00004-of-00004.safetensors",
375
+ "model.layers.29.pre_moe_layernorm.weight": "model-00004-of-00004.safetensors",
376
+ "model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors",
377
+ "model.layers.3.mamba.A_log": "model-00001-of-00004.safetensors",
378
+ "model.layers.3.mamba.B_layernorm.weight": "model-00001-of-00004.safetensors",
379
+ "model.layers.3.mamba.C_layernorm.weight": "model-00001-of-00004.safetensors",
380
+ "model.layers.3.mamba.D": "model-00001-of-00004.safetensors",
381
+ "model.layers.3.mamba.conv1d.bias": "model-00001-of-00004.safetensors",
382
+ "model.layers.3.mamba.conv1d.weight": "model-00001-of-00004.safetensors",
383
+ "model.layers.3.mamba.dt_layernorm.weight": "model-00001-of-00004.safetensors",
384
+ "model.layers.3.mamba.dt_proj.bias": "model-00001-of-00004.safetensors",
385
+ "model.layers.3.mamba.dt_proj.weight": "model-00001-of-00004.safetensors",
386
+ "model.layers.3.mamba.in_proj.weight": "model-00001-of-00004.safetensors",
387
+ "model.layers.3.mamba.out_proj.weight": "model-00001-of-00004.safetensors",
388
+ "model.layers.3.mamba.x_proj.weight": "model-00001-of-00004.safetensors",
389
+ "model.layers.3.moe.experts.0.down_proj.weight": "model-00001-of-00004.safetensors",
390
+ "model.layers.3.moe.experts.0.gate_proj.weight": "model-00001-of-00004.safetensors",
391
+ "model.layers.3.moe.experts.0.up_proj.weight": "model-00001-of-00004.safetensors",
392
+ "model.layers.3.pre_moe_layernorm.weight": "model-00001-of-00004.safetensors",
393
+ "model.layers.30.input_layernorm.weight": "model-00004-of-00004.safetensors",
394
+ "model.layers.30.mamba.A_log": "model-00004-of-00004.safetensors",
395
+ "model.layers.30.mamba.B_layernorm.weight": "model-00004-of-00004.safetensors",
396
+ "model.layers.30.mamba.C_layernorm.weight": "model-00004-of-00004.safetensors",
397
+ "model.layers.30.mamba.D": "model-00004-of-00004.safetensors",
398
+ "model.layers.30.mamba.conv1d.bias": "model-00004-of-00004.safetensors",
399
+ "model.layers.30.mamba.conv1d.weight": "model-00004-of-00004.safetensors",
400
+ "model.layers.30.mamba.dt_layernorm.weight": "model-00004-of-00004.safetensors",
401
+ "model.layers.30.mamba.dt_proj.bias": "model-00004-of-00004.safetensors",
402
+ "model.layers.30.mamba.dt_proj.weight": "model-00004-of-00004.safetensors",
403
+ "model.layers.30.mamba.in_proj.weight": "model-00004-of-00004.safetensors",
404
+ "model.layers.30.mamba.out_proj.weight": "model-00004-of-00004.safetensors",
405
+ "model.layers.30.mamba.x_proj.weight": "model-00004-of-00004.safetensors",
406
+ "model.layers.30.moe.experts.0.down_proj.weight": "model-00004-of-00004.safetensors",
407
+ "model.layers.30.moe.experts.0.gate_proj.weight": "model-00004-of-00004.safetensors",
408
+ "model.layers.30.moe.experts.0.up_proj.weight": "model-00004-of-00004.safetensors",
409
+ "model.layers.30.pre_moe_layernorm.weight": "model-00004-of-00004.safetensors",
410
+ "model.layers.31.input_layernorm.weight": "model-00004-of-00004.safetensors",
411
+ "model.layers.31.mamba.A_log": "model-00004-of-00004.safetensors",
412
+ "model.layers.31.mamba.B_layernorm.weight": "model-00004-of-00004.safetensors",
413
+ "model.layers.31.mamba.C_layernorm.weight": "model-00004-of-00004.safetensors",
414
+ "model.layers.31.mamba.D": "model-00004-of-00004.safetensors",
415
+ "model.layers.31.mamba.conv1d.bias": "model-00004-of-00004.safetensors",
416
+ "model.layers.31.mamba.conv1d.weight": "model-00004-of-00004.safetensors",
417
+ "model.layers.31.mamba.dt_layernorm.weight": "model-00004-of-00004.safetensors",
418
+ "model.layers.31.mamba.dt_proj.bias": "model-00004-of-00004.safetensors",
419
+ "model.layers.31.mamba.dt_proj.weight": "model-00004-of-00004.safetensors",
420
+ "model.layers.31.mamba.in_proj.weight": "model-00004-of-00004.safetensors",
421
+ "model.layers.31.mamba.out_proj.weight": "model-00004-of-00004.safetensors",
422
+ "model.layers.31.mamba.x_proj.weight": "model-00004-of-00004.safetensors",
423
+ "model.layers.31.moe.experts.0.down_proj.weight": "model-00004-of-00004.safetensors",
424
+ "model.layers.31.moe.experts.0.gate_proj.weight": "model-00004-of-00004.safetensors",
425
+ "model.layers.31.moe.experts.0.up_proj.weight": "model-00004-of-00004.safetensors",
426
+ "model.layers.31.pre_moe_layernorm.weight": "model-00004-of-00004.safetensors",
427
+ "model.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors",
428
+ "model.layers.4.moe.experts.0.down_proj.weight": "model-00001-of-00004.safetensors",
429
+ "model.layers.4.moe.experts.0.gate_proj.weight": "model-00001-of-00004.safetensors",
430
+ "model.layers.4.moe.experts.0.up_proj.weight": "model-00001-of-00004.safetensors",
431
+ "model.layers.4.pre_moe_layernorm.weight": "model-00001-of-00004.safetensors",
432
+ "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
433
+ "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
434
+ "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
435
+ "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
436
+ "model.layers.5.input_layernorm.weight": "model-00001-of-00004.safetensors",
437
+ "model.layers.5.mamba.A_log": "model-00001-of-00004.safetensors",
438
+ "model.layers.5.mamba.B_layernorm.weight": "model-00001-of-00004.safetensors",
439
+ "model.layers.5.mamba.C_layernorm.weight": "model-00001-of-00004.safetensors",
440
+ "model.layers.5.mamba.D": "model-00001-of-00004.safetensors",
441
+ "model.layers.5.mamba.conv1d.bias": "model-00001-of-00004.safetensors",
442
+ "model.layers.5.mamba.conv1d.weight": "model-00001-of-00004.safetensors",
443
+ "model.layers.5.mamba.dt_layernorm.weight": "model-00001-of-00004.safetensors",
444
+ "model.layers.5.mamba.dt_proj.bias": "model-00001-of-00004.safetensors",
445
+ "model.layers.5.mamba.dt_proj.weight": "model-00001-of-00004.safetensors",
446
+ "model.layers.5.mamba.in_proj.weight": "model-00001-of-00004.safetensors",
447
+ "model.layers.5.mamba.out_proj.weight": "model-00001-of-00004.safetensors",
448
+ "model.layers.5.mamba.x_proj.weight": "model-00001-of-00004.safetensors",
449
+ "model.layers.5.moe.experts.0.down_proj.weight": "model-00001-of-00004.safetensors",
450
+ "model.layers.5.moe.experts.0.gate_proj.weight": "model-00001-of-00004.safetensors",
451
+ "model.layers.5.moe.experts.0.up_proj.weight": "model-00001-of-00004.safetensors",
452
+ "model.layers.5.pre_moe_layernorm.weight": "model-00001-of-00004.safetensors",
453
+ "model.layers.6.input_layernorm.weight": "model-00001-of-00004.safetensors",
454
+ "model.layers.6.mamba.A_log": "model-00001-of-00004.safetensors",
455
+ "model.layers.6.mamba.B_layernorm.weight": "model-00001-of-00004.safetensors",
456
+ "model.layers.6.mamba.C_layernorm.weight": "model-00001-of-00004.safetensors",
457
+ "model.layers.6.mamba.D": "model-00001-of-00004.safetensors",
458
+ "model.layers.6.mamba.conv1d.bias": "model-00001-of-00004.safetensors",
459
+ "model.layers.6.mamba.conv1d.weight": "model-00001-of-00004.safetensors",
460
+ "model.layers.6.mamba.dt_layernorm.weight": "model-00001-of-00004.safetensors",
461
+ "model.layers.6.mamba.dt_proj.bias": "model-00001-of-00004.safetensors",
462
+ "model.layers.6.mamba.dt_proj.weight": "model-00001-of-00004.safetensors",
463
+ "model.layers.6.mamba.in_proj.weight": "model-00001-of-00004.safetensors",
464
+ "model.layers.6.mamba.out_proj.weight": "model-00001-of-00004.safetensors",
465
+ "model.layers.6.mamba.x_proj.weight": "model-00001-of-00004.safetensors",
466
+ "model.layers.6.moe.experts.0.down_proj.weight": "model-00001-of-00004.safetensors",
467
+ "model.layers.6.moe.experts.0.gate_proj.weight": "model-00001-of-00004.safetensors",
468
+ "model.layers.6.moe.experts.0.up_proj.weight": "model-00001-of-00004.safetensors",
469
+ "model.layers.6.pre_moe_layernorm.weight": "model-00001-of-00004.safetensors",
470
+ "model.layers.7.input_layernorm.weight": "model-00001-of-00004.safetensors",
471
+ "model.layers.7.mamba.A_log": "model-00001-of-00004.safetensors",
472
+ "model.layers.7.mamba.B_layernorm.weight": "model-00001-of-00004.safetensors",
473
+ "model.layers.7.mamba.C_layernorm.weight": "model-00001-of-00004.safetensors",
474
+ "model.layers.7.mamba.D": "model-00001-of-00004.safetensors",
475
+ "model.layers.7.mamba.conv1d.bias": "model-00001-of-00004.safetensors",
476
+ "model.layers.7.mamba.conv1d.weight": "model-00001-of-00004.safetensors",
477
+ "model.layers.7.mamba.dt_layernorm.weight": "model-00001-of-00004.safetensors",
478
+ "model.layers.7.mamba.dt_proj.bias": "model-00001-of-00004.safetensors",
479
+ "model.layers.7.mamba.dt_proj.weight": "model-00001-of-00004.safetensors",
480
+ "model.layers.7.mamba.in_proj.weight": "model-00001-of-00004.safetensors",
481
+ "model.layers.7.mamba.out_proj.weight": "model-00001-of-00004.safetensors",
482
+ "model.layers.7.mamba.x_proj.weight": "model-00001-of-00004.safetensors",
483
+ "model.layers.7.moe.experts.0.down_proj.weight": "model-00001-of-00004.safetensors",
484
+ "model.layers.7.moe.experts.0.gate_proj.weight": "model-00001-of-00004.safetensors",
485
+ "model.layers.7.moe.experts.0.up_proj.weight": "model-00001-of-00004.safetensors",
486
+ "model.layers.7.pre_moe_layernorm.weight": "model-00001-of-00004.safetensors",
487
+ "model.layers.8.input_layernorm.weight": "model-00002-of-00004.safetensors",
488
+ "model.layers.8.mamba.A_log": "model-00001-of-00004.safetensors",
489
+ "model.layers.8.mamba.B_layernorm.weight": "model-00002-of-00004.safetensors",
490
+ "model.layers.8.mamba.C_layernorm.weight": "model-00002-of-00004.safetensors",
491
+ "model.layers.8.mamba.D": "model-00001-of-00004.safetensors",
492
+ "model.layers.8.mamba.conv1d.bias": "model-00001-of-00004.safetensors",
493
+ "model.layers.8.mamba.conv1d.weight": "model-00001-of-00004.safetensors",
494
+ "model.layers.8.mamba.dt_layernorm.weight": "model-00002-of-00004.safetensors",
495
+ "model.layers.8.mamba.dt_proj.bias": "model-00002-of-00004.safetensors",
496
+ "model.layers.8.mamba.dt_proj.weight": "model-00002-of-00004.safetensors",
497
+ "model.layers.8.mamba.in_proj.weight": "model-00002-of-00004.safetensors",
498
+ "model.layers.8.mamba.out_proj.weight": "model-00002-of-00004.safetensors",
499
+ "model.layers.8.mamba.x_proj.weight": "model-00002-of-00004.safetensors",
500
+ "model.layers.8.moe.experts.0.down_proj.weight": "model-00002-of-00004.safetensors",
501
+ "model.layers.8.moe.experts.0.gate_proj.weight": "model-00002-of-00004.safetensors",
502
+ "model.layers.8.moe.experts.0.up_proj.weight": "model-00002-of-00004.safetensors",
503
+ "model.layers.8.pre_moe_layernorm.weight": "model-00002-of-00004.safetensors",
504
+ "model.layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors",
505
+ "model.layers.9.mamba.A_log": "model-00002-of-00004.safetensors",
506
+ "model.layers.9.mamba.B_layernorm.weight": "model-00002-of-00004.safetensors",
507
+ "model.layers.9.mamba.C_layernorm.weight": "model-00002-of-00004.safetensors",
508
+ "model.layers.9.mamba.D": "model-00002-of-00004.safetensors",
509
+ "model.layers.9.mamba.conv1d.bias": "model-00002-of-00004.safetensors",
510
+ "model.layers.9.mamba.conv1d.weight": "model-00002-of-00004.safetensors",
511
+ "model.layers.9.mamba.dt_layernorm.weight": "model-00002-of-00004.safetensors",
512
+ "model.layers.9.mamba.dt_proj.bias": "model-00002-of-00004.safetensors",
513
+ "model.layers.9.mamba.dt_proj.weight": "model-00002-of-00004.safetensors",
514
+ "model.layers.9.mamba.in_proj.weight": "model-00002-of-00004.safetensors",
515
+ "model.layers.9.mamba.out_proj.weight": "model-00002-of-00004.safetensors",
516
+ "model.layers.9.mamba.x_proj.weight": "model-00002-of-00004.safetensors",
517
+ "model.layers.9.moe.experts.0.down_proj.weight": "model-00002-of-00004.safetensors",
518
+ "model.layers.9.moe.experts.0.gate_proj.weight": "model-00002-of-00004.safetensors",
519
+ "model.layers.9.moe.experts.0.up_proj.weight": "model-00002-of-00004.safetensors",
520
+ "model.layers.9.pre_moe_layernorm.weight": "model-00002-of-00004.safetensors"
521
+ }
522
+ }
modeling_jamba.py ADDED
The diff for this file is too large to render. See raw diff
 
special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<|startoftext|>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "<|endoftext|>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "<|pad|>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "<|unk|>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ }
30
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02fd6530b8ede0eedd8e509fcab32da7b1dd04c8119f8498c787100f13112713
3
+ size 1124742
tokenizer_config.json ADDED
@@ -0,0 +1,47 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<|pad|>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<|startoftext|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "<|endoftext|>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "3": {
30
+ "content": "<|unk|>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ }
37
+ },
38
+ "bos_token": "<|startoftext|>",
39
+ "clean_up_tokenization_spaces": false,
40
+ "eos_token": "<|endoftext|>",
41
+ "model_max_length": 1000000000000000019884624838656,
42
+ "pad_token": "<|pad|>",
43
+ "spaces_between_special_tokens": false,
44
+ "tokenizer_class": "LlamaTokenizer",
45
+ "unk_token": "<|unk|>",
46
+ "use_default_system_prompt": false
47
+ }