evanfrick commited on
Commit
3176bf7
·
1 Parent(s): 81117bf

commit files to HF hub

Browse files
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,214 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # lmarena-ai/p2l-7b-bt-01132025
2
+
3
+ Large language model (LLM) evaluations typically rely on aggregated metrics like accuracy or human preference, averaging across users and prompts. This averaging obscures user- and prompt-specific variations in model performance.
4
+ To address this, we propose Prompt-to-Leaderboard (P2L), a method that produces leaderboards specific to a prompt.
5
+ The core idea is to train an LLM taking natural language prompts as input to output a vector of coefficients which are then used to predict the human preference vote.
6
+ The resulting prompt-dependent leaderboards allow for unsupervised task-specific evaluation, optimal routing of queries to models, personalization, and automated evaluation of model strengths and weaknesses.
7
+ Data from Chatbot Arena suggest that P2L better captures the nuanced landscape of language model performance than the averaged leaderboard.
8
+
9
+ **Paper**: [Prompt-to-Leaderboard](https://arxiv.org/abs/2502.14855)
10
+
11
+ **Code**: [lmarena/p2l](https://github.com/lmarena/p2l)
12
+
13
+ This particular P2L model has a *Bradley-Terry* regression head, which we define below:
14
+
15
+ $$
16
+ \begin{equation}
17
+ g_{\theta(z)}(y; x) = \begin{cases}
18
+ \sigma(x^\top \theta^*(z)) & y = 1, \\
19
+ 1-\sigma(x^\top \theta^*(z)) & y = 0.
20
+ \end{cases}
21
+ \end{equation}
22
+ $$
23
+
24
+ More simply, given a prompt, P2L will output a vector of coefficients $\vec{\beta}$. Then the probability that model $i$ beats model $j$, $P(i \succ j) = \sigma(\vec{\beta}_i - \vec{\beta}_j)$.
25
+
26
+ See section 2.2 in our paper for more details on various regression heads.
27
+
28
+ ## Serving
29
+ To serve a P2L model, please see our documentation on GitHub: [Serving P2L](https://github.com/lmarena/p2l?tab=readme-ov-file#serving-p2l).
30
+
31
+ Note: the P2L model outputs with this structure:
32
+
33
+
34
+ ```python
35
+ class P2LOutputs(ModelOutput):
36
+ coefs: torch.FloatTensor = None # "betas" as described above
37
+ eta: Optional[torch.FloatTensor] = None # tie coefficent (not used for BT head)
38
+ last_hidden_state: torch.FloatTensor = None # last hidden state from the transformer
39
+ ```
40
+
41
+ To understand which coefficient index corresponds with which model, see the [`model_list.json`](./model_list.json) found in the repo of each P2L model. As a general rule, the models will always be in sorted order.
42
+
43
+ The easiest way to get this list from inside code is with the following:
44
+
45
+ ```python
46
+ import json
47
+ from huggingface_hub import hf_hub_download
48
+
49
+ fname = hf_hub_download(
50
+ repo_id="lmarena-ai/p2l-7b-bt-01132025", filename="model_list.json", repo_type="model"
51
+ )
52
+
53
+ with open(fname) as fin:
54
+ model_list = json.load(fin)
55
+ ```
56
+
57
+
58
+
59
+ ### Loading from Pretrained
60
+
61
+ To define and load the model:
62
+
63
+ ```python
64
+
65
+ import torch
66
+ from transformers import (
67
+ Qwen2Model,
68
+ Qwen2PreTrainedModel,
69
+ LlamaModel,
70
+ LlamaPreTrainedModel,
71
+ PreTrainedModel,
72
+ AutoTokenizer,
73
+ )
74
+ from transformers import AutoTokenizer
75
+ from transformers.utils import ModelOutput
76
+ from dataclasses import dataclass
77
+ import torch.nn as nn
78
+ import torch.nn.functional as F
79
+ from typing import Dict, Tuple, Callable, Optional
80
+ from huggingface_hub import hf_hub_download
81
+ import json
82
+
83
+
84
+ @dataclass
85
+ class HeadOutputs(ModelOutput):
86
+ coefs: torch.FloatTensor = None
87
+ eta: Optional[torch.FloatTensor] = None
88
+ gamma: Optional[torch.FloatTensor] = None
89
+
90
+
91
+ @dataclass
92
+ class P2LOutputs(ModelOutput):
93
+ coefs: torch.FloatTensor = None
94
+ eta: Optional[torch.FloatTensor] = None
95
+ gamma: Optional[torch.FloatTensor] = None
96
+ loss: Optional[torch.FloatTensor] = None
97
+ last_hidden_state: torch.FloatTensor = None
98
+
99
+ class BTHead(nn.Module):
100
+ def __init__(
101
+ self, input_dim, output_dim, linear_head_downsize_factor=None, **kwargs
102
+ ) -> None:
103
+ super().__init__()
104
+
105
+ if linear_head_downsize_factor:
106
+ inner_dim = int(output_dim // linear_head_downsize_factor)
107
+ self.head = nn.Sequential(
108
+ nn.Linear(in_features=input_dim, out_features=inner_dim, bias=True),
109
+ nn.Linear(in_features=inner_dim, out_features=output_dim, bias=True),
110
+ )
111
+ else:
112
+ self.head = nn.Linear(
113
+ in_features=input_dim, out_features=output_dim, bias=True
114
+ )
115
+
116
+ def forward(self, last_hidden_dim: torch.Tensor):
117
+ coefs = self.head(last_hidden_dim)
118
+ return HeadOutputs(coefs=coefs)
119
+
120
+ class P2LModel(Qwen2PreTrainedModel):
121
+ def __init__(
122
+ self,
123
+ config,
124
+ CLS_id,
125
+ num_models,
126
+ head_kwargs={},
127
+ **kwargs,
128
+ ):
129
+ super().__init__(config)
130
+
131
+ self.num_models = num_models
132
+ self.cls_token_id = CLS_id
133
+
134
+ self.model = Qwen2Model(config)
135
+
136
+ self.head = BTHead(
137
+ input_dim=config.hidden_size,
138
+ output_dim=self.num_models,
139
+ **head_kwargs,
140
+ )
141
+
142
+ self.post_init()
143
+
144
+ def freeze_transformer(self):
145
+ for param in self.model.parameters():
146
+ param.requires_grad = False
147
+
148
+ def get_input_embeddings(self):
149
+ return self.model.embed_tokens
150
+
151
+ def set_input_embeddings(self, value):
152
+ self.model.embed_tokens = value
153
+
154
+ def forward(self, input_ids, attention_mask, labels=None, weights=None):
155
+ batch_size = input_ids.shape[0]
156
+
157
+ hidden_outputs = self.model(
158
+ input_ids=input_ids,
159
+ attention_mask=attention_mask,
160
+ output_hidden_states=False,
161
+ ).last_hidden_state # (bs, num_token, embed_dim)
162
+
163
+ cls_mask = input_ids == self.cls_token_id
164
+
165
+ # double check this is getting the current CLS token
166
+ cls_hidden_dim = hidden_outputs[cls_mask]
167
+
168
+ assert (
169
+ cls_hidden_dim.shape[0] == batch_size
170
+ ), f"input ids {input_ids.shape}, cls_mask {cls_mask.shape}, cls_logit {cls_hidden_dim.shape}"
171
+
172
+ head_output = self.head(cls_hidden_dim)
173
+
174
+
175
+ outputs = P2LOutputs(
176
+ coefs=head_output.coefs,
177
+ last_hidden_state=cls_hidden_dim,
178
+ eta=head_output.eta,
179
+ gamma=head_output.gamma,
180
+ )
181
+
182
+ return outputs
183
+
184
+
185
+ fname = hf_hub_download(
186
+ repo_id="lmarena-ai/p2l-7b-bt-01132025", filename="model_list.json", repo_type="model"
187
+ )
188
+
189
+ with open(fname) as fin:
190
+ model_list = json.load(fin)
191
+
192
+ tokenizer = AutoTokenizer.from_pretrained("lmarena-ai/p2l-7b-bt-01132025")
193
+ model = P2LModel.from_pretrained(
194
+ "lmarena-ai/p2l-7b-bt-01132025",
195
+ CLS_id=tokenizer.cls_token_id,
196
+ num_models=len(model_list),
197
+ torch_dtype=torch.bfloat16,
198
+ )
199
+
200
+ ```
201
+
202
+ ## Citation
203
+
204
+ ```
205
+ @misc{frick2025prompttoleaderboard,
206
+ title={Prompt-to-Leaderboard},
207
+ author={Evan Frick and Connor Chen and Joseph Tennyson and Tianle Li and Wei-Lin Chiang and Anastasios N. Angelopoulos and Ion Stoica},
208
+ year={2025},
209
+ eprint={2502.14855},
210
+ archivePrefix={arXiv},
211
+ primaryClass={cs.LG},
212
+ url={https://arxiv.org/abs/2502.14855},
213
+ }
214
+ ```
added_tokens.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</tool_call>": 151658,
3
+ "<tool_call>": 151657,
4
+ "<|box_end|>": 151649,
5
+ "<|box_start|>": 151648,
6
+ "<|cls|>": 151665,
7
+ "<|endoftext|>": 151643,
8
+ "<|file_sep|>": 151664,
9
+ "<|fim_middle|>": 151660,
10
+ "<|fim_pad|>": 151662,
11
+ "<|fim_prefix|>": 151659,
12
+ "<|fim_suffix|>": 151661,
13
+ "<|im_end|>": 151645,
14
+ "<|im_start|>": 151644,
15
+ "<|image_pad|>": 151655,
16
+ "<|object_ref_end|>": 151647,
17
+ "<|object_ref_start|>": 151646,
18
+ "<|quad_end|>": 151651,
19
+ "<|quad_start|>": 151650,
20
+ "<|repo_name|>": 151663,
21
+ "<|video_pad|>": 151656,
22
+ "<|vision_end|>": 151653,
23
+ "<|vision_pad|>": 151654,
24
+ "<|vision_start|>": 151652
25
+ }
config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
3
+ "architectures": [
4
+ "P2LModel"
5
+ ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 151643,
8
+ "eos_token_id": 151645,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 3584,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 18944,
13
+ "max_position_embeddings": 32768,
14
+ "max_window_layers": 28,
15
+ "model_type": "qwen2",
16
+ "num_attention_heads": 28,
17
+ "num_hidden_layers": 28,
18
+ "num_key_value_heads": 4,
19
+ "rms_norm_eps": 1e-06,
20
+ "rope_scaling": null,
21
+ "rope_theta": 1000000.0,
22
+ "sliding_window": null,
23
+ "tie_word_embeddings": false,
24
+ "torch_dtype": "bfloat16",
25
+ "transformers_version": "4.47.1",
26
+ "use_cache": true,
27
+ "use_sliding_window": false,
28
+ "vocab_size": 152064
29
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd524146a6fde01650bdd3f8982b2264f75cf69f606ce376f2964ffb54826792
3
+ size 4877660776
model-00002-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f16230011678df0c057c321cbf4456e15bfe1ef9db7ebe6a0f116d04b6713698
3
+ size 4932751008
model-00003-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:074568c7b57a492eaebd8c4c6bf0077833da3c9ef421e63b6a1dc1ef7c4d8a39
3
+ size 4331797460
model.safetensors.index.json ADDED
@@ -0,0 +1,347 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 14142170372
4
+ },
5
+ "weight_map": {
6
+ "head.head.bias": "model-00003-of-00003.safetensors",
7
+ "head.head.weight": "model-00003-of-00003.safetensors",
8
+ "model.embed_tokens.weight": "model-00001-of-00003.safetensors",
9
+ "model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors",
10
+ "model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
11
+ "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
12
+ "model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
13
+ "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
14
+ "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
15
+ "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
16
+ "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
17
+ "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
18
+ "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
19
+ "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
20
+ "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
21
+ "model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors",
22
+ "model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
23
+ "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
24
+ "model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
25
+ "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
26
+ "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
27
+ "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
28
+ "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
29
+ "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
30
+ "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
31
+ "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
32
+ "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
33
+ "model.layers.10.input_layernorm.weight": "model-00002-of-00003.safetensors",
34
+ "model.layers.10.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
35
+ "model.layers.10.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
36
+ "model.layers.10.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
37
+ "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
38
+ "model.layers.10.self_attn.k_proj.bias": "model-00002-of-00003.safetensors",
39
+ "model.layers.10.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
40
+ "model.layers.10.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
41
+ "model.layers.10.self_attn.q_proj.bias": "model-00002-of-00003.safetensors",
42
+ "model.layers.10.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
43
+ "model.layers.10.self_attn.v_proj.bias": "model-00002-of-00003.safetensors",
44
+ "model.layers.10.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
45
+ "model.layers.11.input_layernorm.weight": "model-00002-of-00003.safetensors",
46
+ "model.layers.11.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
47
+ "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
48
+ "model.layers.11.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
49
+ "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
50
+ "model.layers.11.self_attn.k_proj.bias": "model-00002-of-00003.safetensors",
51
+ "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
52
+ "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
53
+ "model.layers.11.self_attn.q_proj.bias": "model-00002-of-00003.safetensors",
54
+ "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
55
+ "model.layers.11.self_attn.v_proj.bias": "model-00002-of-00003.safetensors",
56
+ "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
57
+ "model.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors",
58
+ "model.layers.12.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
59
+ "model.layers.12.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
60
+ "model.layers.12.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
61
+ "model.layers.12.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
62
+ "model.layers.12.self_attn.k_proj.bias": "model-00002-of-00003.safetensors",
63
+ "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
64
+ "model.layers.12.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
65
+ "model.layers.12.self_attn.q_proj.bias": "model-00002-of-00003.safetensors",
66
+ "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
67
+ "model.layers.12.self_attn.v_proj.bias": "model-00002-of-00003.safetensors",
68
+ "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
69
+ "model.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors",
70
+ "model.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
71
+ "model.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
72
+ "model.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
73
+ "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
74
+ "model.layers.13.self_attn.k_proj.bias": "model-00002-of-00003.safetensors",
75
+ "model.layers.13.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
76
+ "model.layers.13.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
77
+ "model.layers.13.self_attn.q_proj.bias": "model-00002-of-00003.safetensors",
78
+ "model.layers.13.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
79
+ "model.layers.13.self_attn.v_proj.bias": "model-00002-of-00003.safetensors",
80
+ "model.layers.13.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
81
+ "model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors",
82
+ "model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
83
+ "model.layers.14.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
84
+ "model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
85
+ "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
86
+ "model.layers.14.self_attn.k_proj.bias": "model-00002-of-00003.safetensors",
87
+ "model.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
88
+ "model.layers.14.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
89
+ "model.layers.14.self_attn.q_proj.bias": "model-00002-of-00003.safetensors",
90
+ "model.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
91
+ "model.layers.14.self_attn.v_proj.bias": "model-00002-of-00003.safetensors",
92
+ "model.layers.14.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
93
+ "model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors",
94
+ "model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
95
+ "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
96
+ "model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
97
+ "model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
98
+ "model.layers.15.self_attn.k_proj.bias": "model-00002-of-00003.safetensors",
99
+ "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
100
+ "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
101
+ "model.layers.15.self_attn.q_proj.bias": "model-00002-of-00003.safetensors",
102
+ "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
103
+ "model.layers.15.self_attn.v_proj.bias": "model-00002-of-00003.safetensors",
104
+ "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
105
+ "model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors",
106
+ "model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
107
+ "model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
108
+ "model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
109
+ "model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
110
+ "model.layers.16.self_attn.k_proj.bias": "model-00002-of-00003.safetensors",
111
+ "model.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
112
+ "model.layers.16.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
113
+ "model.layers.16.self_attn.q_proj.bias": "model-00002-of-00003.safetensors",
114
+ "model.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
115
+ "model.layers.16.self_attn.v_proj.bias": "model-00002-of-00003.safetensors",
116
+ "model.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
117
+ "model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors",
118
+ "model.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
119
+ "model.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
120
+ "model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
121
+ "model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
122
+ "model.layers.17.self_attn.k_proj.bias": "model-00002-of-00003.safetensors",
123
+ "model.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
124
+ "model.layers.17.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
125
+ "model.layers.17.self_attn.q_proj.bias": "model-00002-of-00003.safetensors",
126
+ "model.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
127
+ "model.layers.17.self_attn.v_proj.bias": "model-00002-of-00003.safetensors",
128
+ "model.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
129
+ "model.layers.18.input_layernorm.weight": "model-00003-of-00003.safetensors",
130
+ "model.layers.18.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
131
+ "model.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
132
+ "model.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
133
+ "model.layers.18.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
134
+ "model.layers.18.self_attn.k_proj.bias": "model-00002-of-00003.safetensors",
135
+ "model.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
136
+ "model.layers.18.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
137
+ "model.layers.18.self_attn.q_proj.bias": "model-00002-of-00003.safetensors",
138
+ "model.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
139
+ "model.layers.18.self_attn.v_proj.bias": "model-00002-of-00003.safetensors",
140
+ "model.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
141
+ "model.layers.19.input_layernorm.weight": "model-00003-of-00003.safetensors",
142
+ "model.layers.19.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
143
+ "model.layers.19.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
144
+ "model.layers.19.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
145
+ "model.layers.19.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
146
+ "model.layers.19.self_attn.k_proj.bias": "model-00003-of-00003.safetensors",
147
+ "model.layers.19.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
148
+ "model.layers.19.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
149
+ "model.layers.19.self_attn.q_proj.bias": "model-00003-of-00003.safetensors",
150
+ "model.layers.19.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
151
+ "model.layers.19.self_attn.v_proj.bias": "model-00003-of-00003.safetensors",
152
+ "model.layers.19.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
153
+ "model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors",
154
+ "model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
155
+ "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
156
+ "model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
157
+ "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
158
+ "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
159
+ "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
160
+ "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
161
+ "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
162
+ "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
163
+ "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
164
+ "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
165
+ "model.layers.20.input_layernorm.weight": "model-00003-of-00003.safetensors",
166
+ "model.layers.20.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
167
+ "model.layers.20.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
168
+ "model.layers.20.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
169
+ "model.layers.20.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
170
+ "model.layers.20.self_attn.k_proj.bias": "model-00003-of-00003.safetensors",
171
+ "model.layers.20.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
172
+ "model.layers.20.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
173
+ "model.layers.20.self_attn.q_proj.bias": "model-00003-of-00003.safetensors",
174
+ "model.layers.20.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
175
+ "model.layers.20.self_attn.v_proj.bias": "model-00003-of-00003.safetensors",
176
+ "model.layers.20.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
177
+ "model.layers.21.input_layernorm.weight": "model-00003-of-00003.safetensors",
178
+ "model.layers.21.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
179
+ "model.layers.21.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
180
+ "model.layers.21.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
181
+ "model.layers.21.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
182
+ "model.layers.21.self_attn.k_proj.bias": "model-00003-of-00003.safetensors",
183
+ "model.layers.21.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
184
+ "model.layers.21.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
185
+ "model.layers.21.self_attn.q_proj.bias": "model-00003-of-00003.safetensors",
186
+ "model.layers.21.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
187
+ "model.layers.21.self_attn.v_proj.bias": "model-00003-of-00003.safetensors",
188
+ "model.layers.21.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
189
+ "model.layers.22.input_layernorm.weight": "model-00003-of-00003.safetensors",
190
+ "model.layers.22.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
191
+ "model.layers.22.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
192
+ "model.layers.22.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
193
+ "model.layers.22.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
194
+ "model.layers.22.self_attn.k_proj.bias": "model-00003-of-00003.safetensors",
195
+ "model.layers.22.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
196
+ "model.layers.22.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
197
+ "model.layers.22.self_attn.q_proj.bias": "model-00003-of-00003.safetensors",
198
+ "model.layers.22.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
199
+ "model.layers.22.self_attn.v_proj.bias": "model-00003-of-00003.safetensors",
200
+ "model.layers.22.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
201
+ "model.layers.23.input_layernorm.weight": "model-00003-of-00003.safetensors",
202
+ "model.layers.23.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
203
+ "model.layers.23.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
204
+ "model.layers.23.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
205
+ "model.layers.23.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
206
+ "model.layers.23.self_attn.k_proj.bias": "model-00003-of-00003.safetensors",
207
+ "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
208
+ "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
209
+ "model.layers.23.self_attn.q_proj.bias": "model-00003-of-00003.safetensors",
210
+ "model.layers.23.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
211
+ "model.layers.23.self_attn.v_proj.bias": "model-00003-of-00003.safetensors",
212
+ "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
213
+ "model.layers.24.input_layernorm.weight": "model-00003-of-00003.safetensors",
214
+ "model.layers.24.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
215
+ "model.layers.24.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
216
+ "model.layers.24.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
217
+ "model.layers.24.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
218
+ "model.layers.24.self_attn.k_proj.bias": "model-00003-of-00003.safetensors",
219
+ "model.layers.24.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
220
+ "model.layers.24.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
221
+ "model.layers.24.self_attn.q_proj.bias": "model-00003-of-00003.safetensors",
222
+ "model.layers.24.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
223
+ "model.layers.24.self_attn.v_proj.bias": "model-00003-of-00003.safetensors",
224
+ "model.layers.24.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
225
+ "model.layers.25.input_layernorm.weight": "model-00003-of-00003.safetensors",
226
+ "model.layers.25.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
227
+ "model.layers.25.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
228
+ "model.layers.25.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
229
+ "model.layers.25.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
230
+ "model.layers.25.self_attn.k_proj.bias": "model-00003-of-00003.safetensors",
231
+ "model.layers.25.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
232
+ "model.layers.25.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
233
+ "model.layers.25.self_attn.q_proj.bias": "model-00003-of-00003.safetensors",
234
+ "model.layers.25.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
235
+ "model.layers.25.self_attn.v_proj.bias": "model-00003-of-00003.safetensors",
236
+ "model.layers.25.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
237
+ "model.layers.26.input_layernorm.weight": "model-00003-of-00003.safetensors",
238
+ "model.layers.26.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
239
+ "model.layers.26.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
240
+ "model.layers.26.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
241
+ "model.layers.26.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
242
+ "model.layers.26.self_attn.k_proj.bias": "model-00003-of-00003.safetensors",
243
+ "model.layers.26.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
244
+ "model.layers.26.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
245
+ "model.layers.26.self_attn.q_proj.bias": "model-00003-of-00003.safetensors",
246
+ "model.layers.26.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
247
+ "model.layers.26.self_attn.v_proj.bias": "model-00003-of-00003.safetensors",
248
+ "model.layers.26.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
249
+ "model.layers.27.input_layernorm.weight": "model-00003-of-00003.safetensors",
250
+ "model.layers.27.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
251
+ "model.layers.27.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
252
+ "model.layers.27.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
253
+ "model.layers.27.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
254
+ "model.layers.27.self_attn.k_proj.bias": "model-00003-of-00003.safetensors",
255
+ "model.layers.27.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
256
+ "model.layers.27.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
257
+ "model.layers.27.self_attn.q_proj.bias": "model-00003-of-00003.safetensors",
258
+ "model.layers.27.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
259
+ "model.layers.27.self_attn.v_proj.bias": "model-00003-of-00003.safetensors",
260
+ "model.layers.27.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
261
+ "model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors",
262
+ "model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
263
+ "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
264
+ "model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
265
+ "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
266
+ "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
267
+ "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
268
+ "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
269
+ "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
270
+ "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
271
+ "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
272
+ "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
273
+ "model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors",
274
+ "model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
275
+ "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
276
+ "model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
277
+ "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
278
+ "model.layers.4.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
279
+ "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
280
+ "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
281
+ "model.layers.4.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
282
+ "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
283
+ "model.layers.4.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
284
+ "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
285
+ "model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors",
286
+ "model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
287
+ "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
288
+ "model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
289
+ "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
290
+ "model.layers.5.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
291
+ "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
292
+ "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
293
+ "model.layers.5.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
294
+ "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
295
+ "model.layers.5.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
296
+ "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
297
+ "model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors",
298
+ "model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
299
+ "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
300
+ "model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
301
+ "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
302
+ "model.layers.6.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
303
+ "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
304
+ "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
305
+ "model.layers.6.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
306
+ "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
307
+ "model.layers.6.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
308
+ "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
309
+ "model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors",
310
+ "model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
311
+ "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
312
+ "model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
313
+ "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
314
+ "model.layers.7.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
315
+ "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
316
+ "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
317
+ "model.layers.7.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
318
+ "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
319
+ "model.layers.7.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
320
+ "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
321
+ "model.layers.8.input_layernorm.weight": "model-00002-of-00003.safetensors",
322
+ "model.layers.8.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
323
+ "model.layers.8.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
324
+ "model.layers.8.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
325
+ "model.layers.8.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
326
+ "model.layers.8.self_attn.k_proj.bias": "model-00001-of-00003.safetensors",
327
+ "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
328
+ "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
329
+ "model.layers.8.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
330
+ "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
331
+ "model.layers.8.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
332
+ "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
333
+ "model.layers.9.input_layernorm.weight": "model-00002-of-00003.safetensors",
334
+ "model.layers.9.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
335
+ "model.layers.9.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
336
+ "model.layers.9.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
337
+ "model.layers.9.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
338
+ "model.layers.9.self_attn.k_proj.bias": "model-00002-of-00003.safetensors",
339
+ "model.layers.9.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
340
+ "model.layers.9.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
341
+ "model.layers.9.self_attn.q_proj.bias": "model-00002-of-00003.safetensors",
342
+ "model.layers.9.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
343
+ "model.layers.9.self_attn.v_proj.bias": "model-00002-of-00003.safetensors",
344
+ "model.layers.9.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
345
+ "model.norm.weight": "model-00003-of-00003.safetensors"
346
+ }
347
+ }
model_list.json ADDED
@@ -0,0 +1,132 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ "amazon-nova-lite-v1.0",
3
+ "amazon-nova-micro-v1.0",
4
+ "amazon-nova-pro-v1.0",
5
+ "athene-70b-0725",
6
+ "athene-v2-chat",
7
+ "c4ai-aya-expanse-32b",
8
+ "c4ai-aya-expanse-8b",
9
+ "chatgpt-4o-latest-20240808",
10
+ "chatgpt-4o-latest-20240903",
11
+ "chatgpt-4o-latest-20241120",
12
+ "claude-3-5-haiku-20241022",
13
+ "claude-3-5-sonnet-20240620",
14
+ "claude-3-5-sonnet-20241022",
15
+ "claude-3-haiku-20240307",
16
+ "claude-3-opus-20240229",
17
+ "claude-3-sonnet-20240229",
18
+ "codestral-2405",
19
+ "command-r",
20
+ "command-r-08-2024",
21
+ "command-r-plus",
22
+ "command-r-plus-08-2024",
23
+ "dbrx-instruct-preview",
24
+ "deepseek-coder-v2",
25
+ "deepseek-coder-v2-0724",
26
+ "deepseek-v2-api-0628",
27
+ "deepseek-v2.5",
28
+ "deepseek-v2.5-1210",
29
+ "deepseek-v3",
30
+ "gemini-1.5-flash-001",
31
+ "gemini-1.5-flash-002",
32
+ "gemini-1.5-flash-8b-001",
33
+ "gemini-1.5-flash-8b-exp-0827",
34
+ "gemini-1.5-flash-exp-0827",
35
+ "gemini-1.5-pro-001",
36
+ "gemini-1.5-pro-002",
37
+ "gemini-1.5-pro-api-0409-preview",
38
+ "gemini-1.5-pro-exp-0801",
39
+ "gemini-1.5-pro-exp-0827",
40
+ "gemini-2.0-flash-exp",
41
+ "gemini-2.0-flash-thinking-exp-1219",
42
+ "gemini-advanced-0514",
43
+ "gemini-exp-1114",
44
+ "gemini-exp-1121",
45
+ "gemini-exp-1206",
46
+ "gemma-1.1-2b-it",
47
+ "gemma-1.1-7b-it",
48
+ "gemma-2-27b-it",
49
+ "gemma-2-2b-it",
50
+ "gemma-2-9b-it",
51
+ "gemma-2-9b-it-simpo",
52
+ "glm-4-0116",
53
+ "glm-4-0520",
54
+ "glm-4-plus",
55
+ "gpt-3.5-turbo-0125",
56
+ "gpt-4-0125-preview",
57
+ "gpt-4-0314",
58
+ "gpt-4-0613",
59
+ "gpt-4-1106-preview",
60
+ "gpt-4-turbo-2024-04-09",
61
+ "gpt-4o-2024-05-13",
62
+ "gpt-4o-2024-08-06",
63
+ "gpt-4o-mini-2024-07-18",
64
+ "granite-3.0-2b-instruct",
65
+ "granite-3.0-8b-instruct",
66
+ "grok-2-2024-08-13",
67
+ "grok-2-mini-2024-08-13",
68
+ "hunyuan-standard-256k",
69
+ "internlm2_5-20b-chat",
70
+ "jamba-1.5-large",
71
+ "jamba-1.5-mini",
72
+ "llama-2-13b-chat",
73
+ "llama-2-70b-chat",
74
+ "llama-3-70b-instruct",
75
+ "llama-3-8b-instruct",
76
+ "llama-3.1-405b-instruct-bf16",
77
+ "llama-3.1-405b-instruct-fp8",
78
+ "llama-3.1-70b-instruct",
79
+ "llama-3.1-8b-instruct",
80
+ "llama-3.1-nemotron-51b-instruct",
81
+ "llama-3.1-nemotron-70b-instruct",
82
+ "llama-3.1-tulu-3-70b",
83
+ "llama-3.1-tulu-3-8b",
84
+ "llama-3.2-1b-instruct",
85
+ "llama-3.2-3b-instruct",
86
+ "llama-3.3-70b-instruct",
87
+ "ministral-8b-2410",
88
+ "mistral-7b-instruct-v0.2",
89
+ "mistral-large-2402",
90
+ "mistral-large-2407",
91
+ "mistral-large-2411",
92
+ "mistral-medium",
93
+ "mixtral-8x22b-instruct-v0.1",
94
+ "mixtral-8x7b-instruct-v0.1",
95
+ "nemotron-4-340b-instruct",
96
+ "o1-2024-12-17",
97
+ "o1-mini",
98
+ "o1-preview",
99
+ "phi-3-medium-4k-instruct",
100
+ "phi-3-mini-128k-instruct",
101
+ "phi-3-mini-4k-instruct",
102
+ "phi-3-mini-4k-instruct-june-2024",
103
+ "phi-3-small-8k-instruct",
104
+ "qwen-max-0428",
105
+ "qwen-max-0919",
106
+ "qwen-plus-0828",
107
+ "qwen1.5-110b-chat",
108
+ "qwen1.5-14b-chat",
109
+ "qwen1.5-32b-chat",
110
+ "qwen1.5-72b-chat",
111
+ "qwen2-72b-instruct",
112
+ "qwen2.5-72b-instruct",
113
+ "qwen2.5-coder-32b-instruct",
114
+ "qwen2.5-plus-1127",
115
+ "qwq-32b-preview",
116
+ "reka-core-20240501",
117
+ "reka-core-20240722",
118
+ "reka-core-20240904",
119
+ "reka-flash-20240722",
120
+ "reka-flash-20240904",
121
+ "reka-flash-21b-20240226",
122
+ "reka-flash-21b-20240226-online",
123
+ "reka-flash-preview-20240611",
124
+ "smollm2-1.7b-instruct",
125
+ "snowflake-arctic-instruct",
126
+ "yi-1.5-34b-chat",
127
+ "yi-34b-chat",
128
+ "yi-large",
129
+ "yi-large-preview",
130
+ "yi-lightning",
131
+ "yi-lightning-lite"
132
+ ]
special_tokens_map.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>",
5
+ "<|object_ref_start|>",
6
+ "<|object_ref_end|>",
7
+ "<|box_start|>",
8
+ "<|box_end|>",
9
+ "<|quad_start|>",
10
+ "<|quad_end|>",
11
+ "<|vision_start|>",
12
+ "<|vision_end|>",
13
+ "<|vision_pad|>",
14
+ "<|image_pad|>",
15
+ "<|video_pad|>"
16
+ ],
17
+ "cls_token": {
18
+ "content": "<|cls|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ "eos_token": {
25
+ "content": "<|im_end|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ },
31
+ "pad_token": {
32
+ "content": "<|endoftext|>",
33
+ "lstrip": false,
34
+ "normalized": false,
35
+ "rstrip": false,
36
+ "single_word": false
37
+ }
38
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ec927f7a1845161c33d3a6c42ae6aad8e4d597b1fe282ca37b5c794f2498d42
3
+ size 11422346
tokenizer_config.json ADDED
@@ -0,0 +1,217 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_prefix_space": false,
4
+ "added_tokens_decoder": {
5
+ "151643": {
6
+ "content": "<|endoftext|>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "151644": {
14
+ "content": "<|im_start|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "151645": {
22
+ "content": "<|im_end|>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "151646": {
30
+ "content": "<|object_ref_start|>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "151647": {
38
+ "content": "<|object_ref_end|>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": true
44
+ },
45
+ "151648": {
46
+ "content": "<|box_start|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false,
51
+ "special": true
52
+ },
53
+ "151649": {
54
+ "content": "<|box_end|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false,
59
+ "special": true
60
+ },
61
+ "151650": {
62
+ "content": "<|quad_start|>",
63
+ "lstrip": false,
64
+ "normalized": false,
65
+ "rstrip": false,
66
+ "single_word": false,
67
+ "special": true
68
+ },
69
+ "151651": {
70
+ "content": "<|quad_end|>",
71
+ "lstrip": false,
72
+ "normalized": false,
73
+ "rstrip": false,
74
+ "single_word": false,
75
+ "special": true
76
+ },
77
+ "151652": {
78
+ "content": "<|vision_start|>",
79
+ "lstrip": false,
80
+ "normalized": false,
81
+ "rstrip": false,
82
+ "single_word": false,
83
+ "special": true
84
+ },
85
+ "151653": {
86
+ "content": "<|vision_end|>",
87
+ "lstrip": false,
88
+ "normalized": false,
89
+ "rstrip": false,
90
+ "single_word": false,
91
+ "special": true
92
+ },
93
+ "151654": {
94
+ "content": "<|vision_pad|>",
95
+ "lstrip": false,
96
+ "normalized": false,
97
+ "rstrip": false,
98
+ "single_word": false,
99
+ "special": true
100
+ },
101
+ "151655": {
102
+ "content": "<|image_pad|>",
103
+ "lstrip": false,
104
+ "normalized": false,
105
+ "rstrip": false,
106
+ "single_word": false,
107
+ "special": true
108
+ },
109
+ "151656": {
110
+ "content": "<|video_pad|>",
111
+ "lstrip": false,
112
+ "normalized": false,
113
+ "rstrip": false,
114
+ "single_word": false,
115
+ "special": true
116
+ },
117
+ "151657": {
118
+ "content": "<tool_call>",
119
+ "lstrip": false,
120
+ "normalized": false,
121
+ "rstrip": false,
122
+ "single_word": false,
123
+ "special": false
124
+ },
125
+ "151658": {
126
+ "content": "</tool_call>",
127
+ "lstrip": false,
128
+ "normalized": false,
129
+ "rstrip": false,
130
+ "single_word": false,
131
+ "special": false
132
+ },
133
+ "151659": {
134
+ "content": "<|fim_prefix|>",
135
+ "lstrip": false,
136
+ "normalized": false,
137
+ "rstrip": false,
138
+ "single_word": false,
139
+ "special": false
140
+ },
141
+ "151660": {
142
+ "content": "<|fim_middle|>",
143
+ "lstrip": false,
144
+ "normalized": false,
145
+ "rstrip": false,
146
+ "single_word": false,
147
+ "special": false
148
+ },
149
+ "151661": {
150
+ "content": "<|fim_suffix|>",
151
+ "lstrip": false,
152
+ "normalized": false,
153
+ "rstrip": false,
154
+ "single_word": false,
155
+ "special": false
156
+ },
157
+ "151662": {
158
+ "content": "<|fim_pad|>",
159
+ "lstrip": false,
160
+ "normalized": false,
161
+ "rstrip": false,
162
+ "single_word": false,
163
+ "special": false
164
+ },
165
+ "151663": {
166
+ "content": "<|repo_name|>",
167
+ "lstrip": false,
168
+ "normalized": false,
169
+ "rstrip": false,
170
+ "single_word": false,
171
+ "special": false
172
+ },
173
+ "151664": {
174
+ "content": "<|file_sep|>",
175
+ "lstrip": false,
176
+ "normalized": false,
177
+ "rstrip": false,
178
+ "single_word": false,
179
+ "special": false
180
+ },
181
+ "151665": {
182
+ "content": "<|cls|>",
183
+ "lstrip": false,
184
+ "normalized": false,
185
+ "rstrip": false,
186
+ "single_word": false,
187
+ "special": true
188
+ }
189
+ },
190
+ "additional_special_tokens": [
191
+ "<|im_start|>",
192
+ "<|im_end|>",
193
+ "<|object_ref_start|>",
194
+ "<|object_ref_end|>",
195
+ "<|box_start|>",
196
+ "<|box_end|>",
197
+ "<|quad_start|>",
198
+ "<|quad_end|>",
199
+ "<|vision_start|>",
200
+ "<|vision_end|>",
201
+ "<|vision_pad|>",
202
+ "<|image_pad|>",
203
+ "<|video_pad|>"
204
+ ],
205
+ "bos_token": null,
206
+ "chat_template": "{%- if messages[0]['role'] == 'system' %}\n {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n{%- endif %}\n{%- for message in messages %}\n {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n {%- elif message.role == \"assistant\" %}\n {{- '<|im_start|>' + message.role }}\n {%- if message.content %}\n {{- '\\n' + message.content }}\n {%- endif %}\n {%- for tool_call in message.tool_calls %}\n {%- if tool_call.function is defined %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '\\n<tool_call>\\n{\"name\": \"' }}\n {{- tool_call.name }}\n {{- '\", \"arguments\": ' }}\n {{- tool_call.arguments | tojson }}\n {{- '}\\n</tool_call>' }}\n {%- endfor %}\n {{- '<|im_end|>\\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n {{- '<|im_start|>user' }}\n {%- endif %}\n {{- '\\n<tool_response>\\n' }}\n {{- message.content }}\n {{- '\\n</tool_response>' }}\n {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n {{- '<|im_end|>\\n' }}\n {%- endif %}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n\n",
207
+ "clean_up_tokenization_spaces": false,
208
+ "cls_token": "<|cls|>",
209
+ "eos_token": "<|im_end|>",
210
+ "errors": "replace",
211
+ "extra_special_tokens": {},
212
+ "model_max_length": 131072,
213
+ "pad_token": "<|endoftext|>",
214
+ "split_special_tokens": false,
215
+ "tokenizer_class": "Qwen2Tokenizer",
216
+ "unk_token": null
217
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:08ada5a55ecd70a2a913eb31e466d9711da71e575953d3c5be88dc0e061c3d82
3
+ size 7032
training_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "proj_name": "Qwen2.5-7B-Instruct-full-train",
3
+ "learning_rate": 4e-06,
4
+ "adam_epsilon": 1e-08,
5
+ "batch_size": 4,
6
+ "max_length": 8192,
7
+ "num_train_epochs": 1,
8
+ "train_data_path": "full-p2l-data",
9
+ "val_data_path": "p2el/canonical_bt_val_data_11092024",
10
+ "output_dir": "training_outputs",
11
+ "pretrain_model_name": "Qwen/Qwen2.5-7B-Instruct",
12
+ "gradient_accumulation_steps": 16,
13
+ "chat_template": "{%- if messages[0]['role'] == 'system' %}\n {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n{%- endif %}\n{%- for message in messages %}\n {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n {%- elif message.role == \"assistant\" %}\n {{- '<|im_start|>' + message.role }}\n {%- if message.content %}\n {{- '\\n' + message.content }}\n {%- endif %}\n {%- for tool_call in message.tool_calls %}\n {%- if tool_call.function is defined %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '\\n<tool_call>\\n{\"name\": \"' }}\n {{- tool_call.name }}\n {{- '\", \"arguments\": ' }}\n {{- tool_call.arguments | tojson }}\n {{- '}\\n</tool_call>' }}\n {%- endfor %}\n {{- '<|im_end|>\\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n {{- '<|im_start|>user' }}\n {%- endif %}\n {{- '\\n<tool_response>\\n' }}\n {{- message.content }}\n {{- '\\n</tool_response>' }}\n {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n {{- '<|im_end|>\\n' }}\n {%- endif %}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n\n",
14
+ "model_type": "qwen2",
15
+ "head_type": "bt",
16
+ "loss_type": "bt_tie",
17
+ "weighted_loss": false,
18
+ "deepspeed_config_path": "deepspeed/zero1.json",
19
+ "init_type": "reset_params",
20
+ "load_train_data_from_disk": true
21
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff