Commit
•
bb24c22
1
Parent(s):
4b0127f
Upload folder using huggingface_hub
Browse files- README.md +239 -0
- added_tokens.json +5 -0
- config.json +26 -0
- generation_config.json +8 -0
- huggingface-metadata.txt +11 -0
- output.safetensors +3 -0
- pytorch_model.bin.index.json +370 -0
- special_tokens_map.json +24 -0
- tokenizer.model +3 -0
- tokenizer_config.json +37 -0
README.md
ADDED
@@ -0,0 +1,239 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
pipeline_tag: text-generation
|
3 |
+
tags:
|
4 |
+
- orca
|
5 |
+
- orca2
|
6 |
+
- microsoft
|
7 |
+
---
|
8 |
+
|
9 |
+
# Orca 2
|
10 |
+
|
11 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
12 |
+
|
13 |
+
Orca 2 is a helpful assistant that is built for research purposes only and provides a single turn response
|
14 |
+
in tasks such as reasoning over user given data, reading comprehension, math problem solving and text summarization.
|
15 |
+
The model is designed to excel particularly in reasoning.
|
16 |
+
|
17 |
+
We open-source Orca 2 to encourage further research on the development, evaluation, and alignment of smaller LMs.
|
18 |
+
|
19 |
+
## What is Orca 2’s intended use(s)?
|
20 |
+
|
21 |
+
+ Orca 2 is built for research purposes only.
|
22 |
+
+ The main purpose is to allow the research community to assess its abilities and to provide a foundation for
|
23 |
+
building better frontier models.
|
24 |
+
|
25 |
+
## How was Orca 2 evaluated?
|
26 |
+
|
27 |
+
+ Orca 2 has been evaluated on a large number of tasks ranging from reasoning to grounding and safety. Please refer
|
28 |
+
to Section 6 and Appendix in the [Orca 2 paper](https://arxiv.org/pdf/2311.11045.pdf) for details on evaluations.
|
29 |
+
|
30 |
+
## Model Details
|
31 |
+
|
32 |
+
Orca 2 is a finetuned version of LLAMA-2. Orca 2’s training data is a synthetic dataset that was created to enhance the small model’s reasoning abilities.
|
33 |
+
All synthetic training data was moderated using the Microsoft Azure content filters. More details about the model can be found in the [Orca 2 paper](https://arxiv.org/pdf/2311.11045.pdf).
|
34 |
+
|
35 |
+
Please refer to LLaMA-2 technical report for details on the model architecture.
|
36 |
+
|
37 |
+
## License
|
38 |
+
|
39 |
+
Orca 2 is licensed under the [Microsoft Research License](LICENSE).
|
40 |
+
|
41 |
+
Llama 2 is licensed under the [LLAMA 2 Community License](https://ai.meta.com/llama/license/), Copyright © Meta Platforms, Inc. All Rights Reserved.
|
42 |
+
|
43 |
+
## Bias, Risks, and Limitations
|
44 |
+
|
45 |
+
Orca 2, built upon the LLaMA 2 model family, retains many of its limitations, as well as the
|
46 |
+
common limitations of other large language models or limitation caused by its training process,
|
47 |
+
including:
|
48 |
+
|
49 |
+
**Data Biases**: Large language models, trained on extensive data, can inadvertently carry
|
50 |
+
biases present in the source data. Consequently, the models may generate outputs that could
|
51 |
+
be potentially biased or unfair.
|
52 |
+
|
53 |
+
**Lack of Contextual Understanding**: Despite their impressive capabilities in language understanding and generation, these models exhibit limited real-world understanding, resulting
|
54 |
+
in potential inaccuracies or nonsensical responses.
|
55 |
+
|
56 |
+
**Lack of Transparency**: Due to the complexity and size, large language models can act
|
57 |
+
as “black boxes”, making it difficult to comprehend the rationale behind specific outputs or
|
58 |
+
decisions. We recommend reviewing transparency notes from Azure for more information.
|
59 |
+
|
60 |
+
**Content Harms**: There are various types of content harms that large language models
|
61 |
+
can cause. It is important to be aware of them when using these models, and to take
|
62 |
+
actions to prevent them. It is recommended to leverage various content moderation services
|
63 |
+
provided by different companies and institutions. On an important note, we hope for better
|
64 |
+
regulations and standards from government and technology leaders around content harms
|
65 |
+
for AI technologies in future. We value and acknowledge the important role that research
|
66 |
+
and open source community can play in this direction.
|
67 |
+
|
68 |
+
**Hallucination**: It is important to be aware and cautious not to entirely rely on a given
|
69 |
+
language model for critical decisions or information that might have deep impact as it is
|
70 |
+
not obvious how to prevent these models from fabricating content. Moreover, it is not clear
|
71 |
+
whether small models may be more susceptible to hallucination in ungrounded generation
|
72 |
+
use cases due to their smaller sizes and hence reduced memorization capacities. This is an
|
73 |
+
active research topic and we hope there will be more rigorous measurement, understanding
|
74 |
+
and mitigations around this topic.
|
75 |
+
|
76 |
+
**Potential for Misuse**: Without suitable safeguards, there is a risk that these models could
|
77 |
+
be maliciously used for generating disinformation or harmful content.
|
78 |
+
|
79 |
+
**Data Distribution**: Orca 2’s performance is likely to correlate strongly with the distribution
|
80 |
+
of the tuning data. This correlation might limit its accuracy in areas underrepresented in
|
81 |
+
the training dataset such as math, coding, and reasoning.
|
82 |
+
|
83 |
+
**System messages**: Orca 2 demonstrates variance in performance depending on the system
|
84 |
+
instructions. Additionally, the stochasticity introduced by the model size may lead to
|
85 |
+
generation of non-deterministic responses to different system instructions.
|
86 |
+
|
87 |
+
**Zero-Shot Settings**: Orca 2 was trained on data that mostly simulate zero-shot settings.
|
88 |
+
While the model demonstrate very strong performance in zero-shot settings, it does not show
|
89 |
+
the same gains of using few-shot learning compared to other, specially larger, models.
|
90 |
+
|
91 |
+
**Synthetic data**: As Orca 2 is trained on synthetic data, it could inherit both the advantages
|
92 |
+
and shortcomings of the models and methods used for data generation. We posit that Orca
|
93 |
+
2 benefits from the safety measures incorporated during training and safety guardrails (e.g.,
|
94 |
+
content filter) within the Azure OpenAI API. However, detailed studies are required for
|
95 |
+
better quantification of such risks.
|
96 |
+
|
97 |
+
This model is solely designed for research settings, and its testing has only been carried
|
98 |
+
out in such environments. It should not be used in downstream applications, as additional
|
99 |
+
analysis is needed to assess potential harm or bias in the proposed application.
|
100 |
+
|
101 |
+
## Getting started with Orca 2
|
102 |
+
|
103 |
+
**Inference with Hugging Face library**
|
104 |
+
|
105 |
+
```python
|
106 |
+
import torch
|
107 |
+
import transformers
|
108 |
+
|
109 |
+
if torch.cuda.is_available():
|
110 |
+
torch.set_default_device("cuda")
|
111 |
+
else:
|
112 |
+
torch.set_default_device("cpu")
|
113 |
+
|
114 |
+
model = transformers.AutoModelForCausalLM.from_pretrained("microsoft/Orca-2-13b", device_map='auto')
|
115 |
+
|
116 |
+
# https://github.com/huggingface/transformers/issues/27132
|
117 |
+
# please use the slow tokenizer since fast and slow tokenizer produces different tokens
|
118 |
+
tokenizer = transformers.AutoTokenizer.from_pretrained(
|
119 |
+
"microsoft/Orca-2-13b",
|
120 |
+
use_fast=False,
|
121 |
+
)
|
122 |
+
|
123 |
+
system_message = "You are Orca, an AI language model created by Microsoft. You are a cautious assistant. You carefully follow instructions. You are helpful and harmless and you follow ethical guidelines and promote positive behavior."
|
124 |
+
user_message = "How can you determine if a restaurant is popular among locals or mainly attracts tourists, and why might this information be useful?"
|
125 |
+
|
126 |
+
prompt = f"<|im_start|>system\n{system_message}<|im_end|>\n<|im_start|>user\n{user_message}<|im_end|>\n<|im_start|>assistant"
|
127 |
+
|
128 |
+
inputs = tokenizer(prompt, return_tensors='pt')
|
129 |
+
output_ids = model.generate(inputs["input_ids"],)
|
130 |
+
answer = tokenizer.batch_decode(output_ids)[0]
|
131 |
+
|
132 |
+
print(answer)
|
133 |
+
|
134 |
+
# This example continues showing how to add a second turn message by the user to the conversation
|
135 |
+
second_turn_user_message = "Give me a list of the key points of your first answer."
|
136 |
+
|
137 |
+
# we set add_special_tokens=False because we dont want to automatically add a bos_token between messages
|
138 |
+
second_turn_message_in_markup = f"\n<|im_start|>user\n{second_turn_user_message}<|im_end|>\n<|im_start|>assistant"
|
139 |
+
second_turn_tokens = tokenizer(second_turn_message_in_markup, return_tensors='pt', add_special_tokens=False)
|
140 |
+
second_turn_input = torch.cat([output_ids, second_turn_tokens['input_ids']], dim=1)
|
141 |
+
|
142 |
+
output_ids_2 = model.generate(second_turn_input,)
|
143 |
+
second_turn_answer = tokenizer.batch_decode(output_ids_2)[0]
|
144 |
+
|
145 |
+
print(second_turn_answer)
|
146 |
+
```
|
147 |
+
|
148 |
+
|
149 |
+
**Safe inference with Azure AI Content Safety**
|
150 |
+
|
151 |
+
The usage of [Azure AI Content Safety](https://azure.microsoft.com/en-us/products/ai-services/ai-content-safety/) on top of model prediction is strongly encouraged
|
152 |
+
and can help prevent content harms. Azure AI Content Safety is a content moderation platform
|
153 |
+
that uses AI to keep your content safe. By integrating Orca 2 with Azure AI Content Safety,
|
154 |
+
we can moderate the model output by scanning it for sexual content, violence, hate, and
|
155 |
+
self-harm with multiple severity levels and multi-lingual detection.
|
156 |
+
|
157 |
+
```python
|
158 |
+
import os
|
159 |
+
import math
|
160 |
+
import transformers
|
161 |
+
import torch
|
162 |
+
|
163 |
+
from azure.ai.contentsafety import ContentSafetyClient
|
164 |
+
from azure.core.credentials import AzureKeyCredential
|
165 |
+
from azure.core.exceptions import HttpResponseError
|
166 |
+
from azure.ai.contentsafety.models import AnalyzeTextOptions
|
167 |
+
|
168 |
+
CONTENT_SAFETY_KEY = os.environ["CONTENT_SAFETY_KEY"]
|
169 |
+
CONTENT_SAFETY_ENDPOINT = os.environ["CONTENT_SAFETY_ENDPOINT"]
|
170 |
+
|
171 |
+
# We use Azure AI Content Safety to filter out any content that reaches "Medium" threshold
|
172 |
+
# For more information: https://learn.microsoft.com/en-us/azure/ai-services/content-safety/
|
173 |
+
def should_filter_out(input_text, threshold=4):
|
174 |
+
# Create an Content Safety client
|
175 |
+
client = ContentSafetyClient(CONTENT_SAFETY_ENDPOINT, AzureKeyCredential(CONTENT_SAFETY_KEY))
|
176 |
+
|
177 |
+
# Construct a request
|
178 |
+
request = AnalyzeTextOptions(text=input_text)
|
179 |
+
|
180 |
+
# Analyze text
|
181 |
+
try:
|
182 |
+
response = client.analyze_text(request)
|
183 |
+
except HttpResponseError as e:
|
184 |
+
print("Analyze text failed.")
|
185 |
+
if e.error:
|
186 |
+
print(f"Error code: {e.error.code}")
|
187 |
+
print(f"Error message: {e.error.message}")
|
188 |
+
raise
|
189 |
+
print(e)
|
190 |
+
raise
|
191 |
+
|
192 |
+
categories = ["hate_result", "self_harm_result", "sexual_result", "violence_result"]
|
193 |
+
max_score = -math.inf
|
194 |
+
for category in categories:
|
195 |
+
max_score = max(max_score, getattr(response, category).severity)
|
196 |
+
|
197 |
+
return max_score >= threshold
|
198 |
+
|
199 |
+
model_path = 'microsoft/Orca-2-13b'
|
200 |
+
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
|
201 |
+
model = transformers.AutoModelForCausalLM.from_pretrained(model_path)
|
202 |
+
model.to(device)
|
203 |
+
|
204 |
+
tokenizer = transformers.AutoTokenizer.from_pretrained(
|
205 |
+
model_path,
|
206 |
+
model_max_length=4096,
|
207 |
+
padding_side="right",
|
208 |
+
use_fast=False,
|
209 |
+
add_special_tokens=False,
|
210 |
+
)
|
211 |
+
|
212 |
+
system_message = "You are Orca, an AI language model created by Microsoft. You are a cautious assistant. You carefully follow instructions. You are helpful and harmless and you follow ethical guidelines and promote positive behavior."
|
213 |
+
user_message = "\" \n :You can't just say, \"\"that's crap\"\" and remove it without gaining a consensus. You already know this, based on your block history. —/ \" \nIs the comment obscene? \nOptions : Yes, No."
|
214 |
+
|
215 |
+
prompt = f"<|im_start|>system\n{system_message}<|im_end|>\n<|im_start|>user\n{user_message}<|im_end|>\n<|im_start|>assistant"
|
216 |
+
|
217 |
+
inputs = tokenizer(prompt, return_tensors='pt')
|
218 |
+
inputs = inputs.to(device)
|
219 |
+
|
220 |
+
output_ids = model.generate(inputs["input_ids"], max_length=4096, do_sample=False, temperature=0.0, use_cache=True)
|
221 |
+
sequence_length = inputs["input_ids"].shape[1]
|
222 |
+
new_output_ids = output_ids[:, sequence_length:]
|
223 |
+
answers = tokenizer.batch_decode(new_output_ids, skip_special_tokens=True)
|
224 |
+
final_output = answers[0] if not should_filter_out(answers[0]) else "[Content Filtered]"
|
225 |
+
|
226 |
+
print(final_output)
|
227 |
+
```
|
228 |
+
|
229 |
+
## Citation
|
230 |
+
```bibtex
|
231 |
+
@misc{mitra2023orca,
|
232 |
+
title={Orca 2: Teaching Small Language Models How to Reason},
|
233 |
+
author={Arindam Mitra and Luciano Del Corro and Shweti Mahajan and Andres Codas and Clarisse Simoes and Sahaj Agrawal and Xuxi Chen and Anastasia Razdaibiedina and Erik Jones and Kriti Aggarwal and Hamid Palangi and Guoqing Zheng and Corby Rosset and Hamed Khanpour and Ahmed Awadallah},
|
234 |
+
year={2023},
|
235 |
+
eprint={2311.11045},
|
236 |
+
archivePrefix={arXiv},
|
237 |
+
primaryClass={cs.AI}
|
238 |
+
}
|
239 |
+
```
|
added_tokens.json
ADDED
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"<|im_end|>": 32002,
|
3 |
+
"<|im_start|>": 32001,
|
4 |
+
"[PAD]": 32000
|
5 |
+
}
|
config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"architectures": [
|
3 |
+
"LlamaForCausalLM"
|
4 |
+
],
|
5 |
+
"bos_token_id": 1,
|
6 |
+
"eos_token_id": 2,
|
7 |
+
"hidden_act": "silu",
|
8 |
+
"hidden_size": 5120,
|
9 |
+
"initializer_range": 0.02,
|
10 |
+
"intermediate_size": 13824,
|
11 |
+
"max_position_embeddings": 4096,
|
12 |
+
"model_type": "llama",
|
13 |
+
"num_attention_heads": 40,
|
14 |
+
"num_hidden_layers": 40,
|
15 |
+
"num_key_value_heads": 40,
|
16 |
+
"pad_token_id": 0,
|
17 |
+
"pretraining_tp": 1,
|
18 |
+
"rms_norm_eps": 1e-05,
|
19 |
+
"rope_scaling": null,
|
20 |
+
"rope_theta": 10000.0,
|
21 |
+
"tie_word_embeddings": false,
|
22 |
+
"torch_dtype": "float32",
|
23 |
+
"transformers_version": "4.33.1",
|
24 |
+
"use_cache": true,
|
25 |
+
"vocab_size": 32003
|
26 |
+
}
|
generation_config.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token_id": 1,
|
3 |
+
"do_sample": false,
|
4 |
+
"eos_token_id": 2,
|
5 |
+
"max_length": 4096,
|
6 |
+
"pad_token_id": 0,
|
7 |
+
"transformers_version": "4.33.1"
|
8 |
+
}
|
huggingface-metadata.txt
ADDED
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
url: https://huggingface.co/microsoft/Orca-2-13b
|
2 |
+
branch: main
|
3 |
+
download date: 2023-11-20 21:44:37
|
4 |
+
sha256sum:
|
5 |
+
48b50d9d6ef77522b210c62ec214dd01295fcf88a3160acce32cae23d033ce1d pytorch_model-00001-of-00006.bin
|
6 |
+
d5dec51a499eea69793a89ce4409b019092829a8efbe5ca0f73cd1da3b7431d3 pytorch_model-00002-of-00006.bin
|
7 |
+
5bd111371970c260e3b38685f6a39bc54ba189e9aa1297e142b9193e21fab415 pytorch_model-00003-of-00006.bin
|
8 |
+
9870c5705b3b409ba7ee8c60c6e785c350d8f69194ea58db11a8acfef808a4cf pytorch_model-00004-of-00006.bin
|
9 |
+
e17d513a94757fed6a86e48d9f425fcf0a52f8bc47b83f84e809eabc5e685196 pytorch_model-00005-of-00006.bin
|
10 |
+
974119123f7033984e573b0d2c9754275bd448c688a09488e52f2f44206785e1 pytorch_model-00006-of-00006.bin
|
11 |
+
9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347 tokenizer.model
|
output.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:84d980e42df837c166d03a843ae2ae6221ff104dad4ebc71be57cede8b9ee905
|
3 |
+
size 5225400600
|
pytorch_model.bin.index.json
ADDED
@@ -0,0 +1,370 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"metadata": {
|
3 |
+
"total_size": 52063580160
|
4 |
+
},
|
5 |
+
"weight_map": {
|
6 |
+
"lm_head.weight": "pytorch_model-00006-of-00006.bin",
|
7 |
+
"model.embed_tokens.weight": "pytorch_model-00001-of-00006.bin",
|
8 |
+
"model.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
9 |
+
"model.layers.0.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
|
10 |
+
"model.layers.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
|
11 |
+
"model.layers.0.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
|
12 |
+
"model.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
13 |
+
"model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
|
14 |
+
"model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
|
15 |
+
"model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
|
16 |
+
"model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
|
17 |
+
"model.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
18 |
+
"model.layers.1.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
|
19 |
+
"model.layers.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
|
20 |
+
"model.layers.1.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
|
21 |
+
"model.layers.1.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
22 |
+
"model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
|
23 |
+
"model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
|
24 |
+
"model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
|
25 |
+
"model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
|
26 |
+
"model.layers.10.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
27 |
+
"model.layers.10.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
|
28 |
+
"model.layers.10.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
|
29 |
+
"model.layers.10.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
|
30 |
+
"model.layers.10.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
31 |
+
"model.layers.10.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
|
32 |
+
"model.layers.10.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
|
33 |
+
"model.layers.10.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
|
34 |
+
"model.layers.10.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
|
35 |
+
"model.layers.11.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
36 |
+
"model.layers.11.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
|
37 |
+
"model.layers.11.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
|
38 |
+
"model.layers.11.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
|
39 |
+
"model.layers.11.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
40 |
+
"model.layers.11.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
|
41 |
+
"model.layers.11.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
|
42 |
+
"model.layers.11.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
|
43 |
+
"model.layers.11.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
|
44 |
+
"model.layers.12.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
45 |
+
"model.layers.12.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
|
46 |
+
"model.layers.12.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
|
47 |
+
"model.layers.12.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
|
48 |
+
"model.layers.12.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
49 |
+
"model.layers.12.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
|
50 |
+
"model.layers.12.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
|
51 |
+
"model.layers.12.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
|
52 |
+
"model.layers.12.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
|
53 |
+
"model.layers.13.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
54 |
+
"model.layers.13.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
|
55 |
+
"model.layers.13.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
|
56 |
+
"model.layers.13.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
|
57 |
+
"model.layers.13.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
58 |
+
"model.layers.13.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
|
59 |
+
"model.layers.13.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
|
60 |
+
"model.layers.13.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
|
61 |
+
"model.layers.13.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
|
62 |
+
"model.layers.14.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
63 |
+
"model.layers.14.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
|
64 |
+
"model.layers.14.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
|
65 |
+
"model.layers.14.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
|
66 |
+
"model.layers.14.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
67 |
+
"model.layers.14.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
|
68 |
+
"model.layers.14.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
|
69 |
+
"model.layers.14.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
|
70 |
+
"model.layers.14.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
|
71 |
+
"model.layers.15.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
72 |
+
"model.layers.15.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
|
73 |
+
"model.layers.15.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
|
74 |
+
"model.layers.15.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
|
75 |
+
"model.layers.15.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
76 |
+
"model.layers.15.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
|
77 |
+
"model.layers.15.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
|
78 |
+
"model.layers.15.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
|
79 |
+
"model.layers.15.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
|
80 |
+
"model.layers.16.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
81 |
+
"model.layers.16.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
|
82 |
+
"model.layers.16.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
|
83 |
+
"model.layers.16.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
|
84 |
+
"model.layers.16.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
85 |
+
"model.layers.16.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
|
86 |
+
"model.layers.16.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
|
87 |
+
"model.layers.16.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
|
88 |
+
"model.layers.16.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
|
89 |
+
"model.layers.17.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
90 |
+
"model.layers.17.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
|
91 |
+
"model.layers.17.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
|
92 |
+
"model.layers.17.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
|
93 |
+
"model.layers.17.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
94 |
+
"model.layers.17.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
|
95 |
+
"model.layers.17.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
|
96 |
+
"model.layers.17.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
|
97 |
+
"model.layers.17.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
|
98 |
+
"model.layers.18.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
99 |
+
"model.layers.18.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
|
100 |
+
"model.layers.18.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
|
101 |
+
"model.layers.18.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
|
102 |
+
"model.layers.18.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
103 |
+
"model.layers.18.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
|
104 |
+
"model.layers.18.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
|
105 |
+
"model.layers.18.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
|
106 |
+
"model.layers.18.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
|
107 |
+
"model.layers.19.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
108 |
+
"model.layers.19.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
|
109 |
+
"model.layers.19.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
|
110 |
+
"model.layers.19.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
|
111 |
+
"model.layers.19.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
112 |
+
"model.layers.19.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
|
113 |
+
"model.layers.19.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
|
114 |
+
"model.layers.19.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
|
115 |
+
"model.layers.19.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
|
116 |
+
"model.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
117 |
+
"model.layers.2.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
|
118 |
+
"model.layers.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
|
119 |
+
"model.layers.2.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
|
120 |
+
"model.layers.2.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
121 |
+
"model.layers.2.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
|
122 |
+
"model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
|
123 |
+
"model.layers.2.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
|
124 |
+
"model.layers.2.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
|
125 |
+
"model.layers.20.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
126 |
+
"model.layers.20.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
|
127 |
+
"model.layers.20.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
|
128 |
+
"model.layers.20.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
|
129 |
+
"model.layers.20.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
130 |
+
"model.layers.20.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
|
131 |
+
"model.layers.20.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
|
132 |
+
"model.layers.20.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
|
133 |
+
"model.layers.20.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
|
134 |
+
"model.layers.21.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
135 |
+
"model.layers.21.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
|
136 |
+
"model.layers.21.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
|
137 |
+
"model.layers.21.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
|
138 |
+
"model.layers.21.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
139 |
+
"model.layers.21.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
|
140 |
+
"model.layers.21.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
|
141 |
+
"model.layers.21.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
|
142 |
+
"model.layers.21.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
|
143 |
+
"model.layers.22.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
144 |
+
"model.layers.22.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
|
145 |
+
"model.layers.22.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
|
146 |
+
"model.layers.22.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
|
147 |
+
"model.layers.22.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
|
148 |
+
"model.layers.22.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
|
149 |
+
"model.layers.22.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
|
150 |
+
"model.layers.22.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
|
151 |
+
"model.layers.22.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
|
152 |
+
"model.layers.23.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
153 |
+
"model.layers.23.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
|
154 |
+
"model.layers.23.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
|
155 |
+
"model.layers.23.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
|
156 |
+
"model.layers.23.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
157 |
+
"model.layers.23.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
|
158 |
+
"model.layers.23.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
|
159 |
+
"model.layers.23.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
|
160 |
+
"model.layers.23.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
|
161 |
+
"model.layers.24.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
162 |
+
"model.layers.24.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
|
163 |
+
"model.layers.24.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
|
164 |
+
"model.layers.24.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
|
165 |
+
"model.layers.24.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
166 |
+
"model.layers.24.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
|
167 |
+
"model.layers.24.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
|
168 |
+
"model.layers.24.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
|
169 |
+
"model.layers.24.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
|
170 |
+
"model.layers.25.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
171 |
+
"model.layers.25.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
|
172 |
+
"model.layers.25.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
|
173 |
+
"model.layers.25.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
|
174 |
+
"model.layers.25.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
175 |
+
"model.layers.25.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
|
176 |
+
"model.layers.25.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
|
177 |
+
"model.layers.25.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
|
178 |
+
"model.layers.25.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
|
179 |
+
"model.layers.26.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
180 |
+
"model.layers.26.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
|
181 |
+
"model.layers.26.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
|
182 |
+
"model.layers.26.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
|
183 |
+
"model.layers.26.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
184 |
+
"model.layers.26.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
|
185 |
+
"model.layers.26.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
|
186 |
+
"model.layers.26.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
|
187 |
+
"model.layers.26.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
|
188 |
+
"model.layers.27.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
189 |
+
"model.layers.27.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
|
190 |
+
"model.layers.27.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
|
191 |
+
"model.layers.27.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
|
192 |
+
"model.layers.27.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
193 |
+
"model.layers.27.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
|
194 |
+
"model.layers.27.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
|
195 |
+
"model.layers.27.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
|
196 |
+
"model.layers.27.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
|
197 |
+
"model.layers.28.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
198 |
+
"model.layers.28.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
|
199 |
+
"model.layers.28.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
|
200 |
+
"model.layers.28.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
|
201 |
+
"model.layers.28.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
202 |
+
"model.layers.28.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
|
203 |
+
"model.layers.28.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
|
204 |
+
"model.layers.28.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
|
205 |
+
"model.layers.28.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
|
206 |
+
"model.layers.29.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
207 |
+
"model.layers.29.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
|
208 |
+
"model.layers.29.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
|
209 |
+
"model.layers.29.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
|
210 |
+
"model.layers.29.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
|
211 |
+
"model.layers.29.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
|
212 |
+
"model.layers.29.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
|
213 |
+
"model.layers.29.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
|
214 |
+
"model.layers.29.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
|
215 |
+
"model.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
216 |
+
"model.layers.3.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
|
217 |
+
"model.layers.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
|
218 |
+
"model.layers.3.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
|
219 |
+
"model.layers.3.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
220 |
+
"model.layers.3.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
|
221 |
+
"model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
|
222 |
+
"model.layers.3.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
|
223 |
+
"model.layers.3.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
|
224 |
+
"model.layers.30.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
225 |
+
"model.layers.30.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
|
226 |
+
"model.layers.30.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
|
227 |
+
"model.layers.30.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
|
228 |
+
"model.layers.30.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
229 |
+
"model.layers.30.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
|
230 |
+
"model.layers.30.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
|
231 |
+
"model.layers.30.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
|
232 |
+
"model.layers.30.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
|
233 |
+
"model.layers.31.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
234 |
+
"model.layers.31.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
|
235 |
+
"model.layers.31.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
|
236 |
+
"model.layers.31.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
|
237 |
+
"model.layers.31.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
238 |
+
"model.layers.31.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
|
239 |
+
"model.layers.31.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
|
240 |
+
"model.layers.31.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
|
241 |
+
"model.layers.31.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
|
242 |
+
"model.layers.32.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
243 |
+
"model.layers.32.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
|
244 |
+
"model.layers.32.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
|
245 |
+
"model.layers.32.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
|
246 |
+
"model.layers.32.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
247 |
+
"model.layers.32.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
|
248 |
+
"model.layers.32.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
|
249 |
+
"model.layers.32.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
|
250 |
+
"model.layers.32.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
|
251 |
+
"model.layers.33.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
252 |
+
"model.layers.33.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
|
253 |
+
"model.layers.33.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
|
254 |
+
"model.layers.33.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
|
255 |
+
"model.layers.33.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
256 |
+
"model.layers.33.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
|
257 |
+
"model.layers.33.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
|
258 |
+
"model.layers.33.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
|
259 |
+
"model.layers.33.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
|
260 |
+
"model.layers.34.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
261 |
+
"model.layers.34.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
|
262 |
+
"model.layers.34.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
|
263 |
+
"model.layers.34.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
|
264 |
+
"model.layers.34.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
265 |
+
"model.layers.34.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
|
266 |
+
"model.layers.34.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
|
267 |
+
"model.layers.34.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
|
268 |
+
"model.layers.34.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
|
269 |
+
"model.layers.35.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
270 |
+
"model.layers.35.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
|
271 |
+
"model.layers.35.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
|
272 |
+
"model.layers.35.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
|
273 |
+
"model.layers.35.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
274 |
+
"model.layers.35.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
|
275 |
+
"model.layers.35.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
|
276 |
+
"model.layers.35.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
|
277 |
+
"model.layers.35.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
|
278 |
+
"model.layers.36.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
279 |
+
"model.layers.36.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
|
280 |
+
"model.layers.36.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
|
281 |
+
"model.layers.36.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
|
282 |
+
"model.layers.36.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
283 |
+
"model.layers.36.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
|
284 |
+
"model.layers.36.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
|
285 |
+
"model.layers.36.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
|
286 |
+
"model.layers.36.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
|
287 |
+
"model.layers.37.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
288 |
+
"model.layers.37.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
|
289 |
+
"model.layers.37.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
|
290 |
+
"model.layers.37.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
|
291 |
+
"model.layers.37.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
|
292 |
+
"model.layers.37.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
|
293 |
+
"model.layers.37.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
|
294 |
+
"model.layers.37.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
|
295 |
+
"model.layers.37.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
|
296 |
+
"model.layers.38.input_layernorm.weight": "pytorch_model-00006-of-00006.bin",
|
297 |
+
"model.layers.38.mlp.down_proj.weight": "pytorch_model-00006-of-00006.bin",
|
298 |
+
"model.layers.38.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
|
299 |
+
"model.layers.38.mlp.up_proj.weight": "pytorch_model-00006-of-00006.bin",
|
300 |
+
"model.layers.38.post_attention_layernorm.weight": "pytorch_model-00006-of-00006.bin",
|
301 |
+
"model.layers.38.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
|
302 |
+
"model.layers.38.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
|
303 |
+
"model.layers.38.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
|
304 |
+
"model.layers.38.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
|
305 |
+
"model.layers.39.input_layernorm.weight": "pytorch_model-00006-of-00006.bin",
|
306 |
+
"model.layers.39.mlp.down_proj.weight": "pytorch_model-00006-of-00006.bin",
|
307 |
+
"model.layers.39.mlp.gate_proj.weight": "pytorch_model-00006-of-00006.bin",
|
308 |
+
"model.layers.39.mlp.up_proj.weight": "pytorch_model-00006-of-00006.bin",
|
309 |
+
"model.layers.39.post_attention_layernorm.weight": "pytorch_model-00006-of-00006.bin",
|
310 |
+
"model.layers.39.self_attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
|
311 |
+
"model.layers.39.self_attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
|
312 |
+
"model.layers.39.self_attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
|
313 |
+
"model.layers.39.self_attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
|
314 |
+
"model.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
315 |
+
"model.layers.4.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
|
316 |
+
"model.layers.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
|
317 |
+
"model.layers.4.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
|
318 |
+
"model.layers.4.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
319 |
+
"model.layers.4.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
|
320 |
+
"model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
|
321 |
+
"model.layers.4.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
|
322 |
+
"model.layers.4.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
|
323 |
+
"model.layers.5.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
324 |
+
"model.layers.5.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
|
325 |
+
"model.layers.5.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
|
326 |
+
"model.layers.5.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
|
327 |
+
"model.layers.5.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
328 |
+
"model.layers.5.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
|
329 |
+
"model.layers.5.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
|
330 |
+
"model.layers.5.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
|
331 |
+
"model.layers.5.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
|
332 |
+
"model.layers.6.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
333 |
+
"model.layers.6.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
|
334 |
+
"model.layers.6.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
|
335 |
+
"model.layers.6.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
|
336 |
+
"model.layers.6.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
|
337 |
+
"model.layers.6.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
|
338 |
+
"model.layers.6.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
|
339 |
+
"model.layers.6.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
|
340 |
+
"model.layers.6.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
|
341 |
+
"model.layers.7.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
342 |
+
"model.layers.7.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
|
343 |
+
"model.layers.7.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
|
344 |
+
"model.layers.7.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
|
345 |
+
"model.layers.7.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
346 |
+
"model.layers.7.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
|
347 |
+
"model.layers.7.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
|
348 |
+
"model.layers.7.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
|
349 |
+
"model.layers.7.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
|
350 |
+
"model.layers.8.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
351 |
+
"model.layers.8.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
|
352 |
+
"model.layers.8.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
|
353 |
+
"model.layers.8.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
|
354 |
+
"model.layers.8.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
355 |
+
"model.layers.8.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
|
356 |
+
"model.layers.8.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
|
357 |
+
"model.layers.8.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
|
358 |
+
"model.layers.8.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
|
359 |
+
"model.layers.9.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
360 |
+
"model.layers.9.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
|
361 |
+
"model.layers.9.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
|
362 |
+
"model.layers.9.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
|
363 |
+
"model.layers.9.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
|
364 |
+
"model.layers.9.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
|
365 |
+
"model.layers.9.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
|
366 |
+
"model.layers.9.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
|
367 |
+
"model.layers.9.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
|
368 |
+
"model.norm.weight": "pytorch_model-00006-of-00006.bin"
|
369 |
+
}
|
370 |
+
}
|
special_tokens_map.json
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": {
|
3 |
+
"content": "<s>",
|
4 |
+
"lstrip": false,
|
5 |
+
"normalized": false,
|
6 |
+
"rstrip": false,
|
7 |
+
"single_word": false
|
8 |
+
},
|
9 |
+
"eos_token": {
|
10 |
+
"content": "</s>",
|
11 |
+
"lstrip": false,
|
12 |
+
"normalized": false,
|
13 |
+
"rstrip": false,
|
14 |
+
"single_word": false
|
15 |
+
},
|
16 |
+
"pad_token": "[PAD]",
|
17 |
+
"unk_token": {
|
18 |
+
"content": "<unk>",
|
19 |
+
"lstrip": false,
|
20 |
+
"normalized": false,
|
21 |
+
"rstrip": false,
|
22 |
+
"single_word": false
|
23 |
+
}
|
24 |
+
}
|
tokenizer.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
3 |
+
size 499723
|
tokenizer_config.json
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"add_bos_token": true,
|
3 |
+
"add_eos_token": false,
|
4 |
+
"bos_token": {
|
5 |
+
"__type": "AddedToken",
|
6 |
+
"content": "<s>",
|
7 |
+
"lstrip": false,
|
8 |
+
"normalized": false,
|
9 |
+
"rstrip": false,
|
10 |
+
"single_word": false
|
11 |
+
},
|
12 |
+
"clean_up_tokenization_spaces": false,
|
13 |
+
"eos_token": {
|
14 |
+
"__type": "AddedToken",
|
15 |
+
"content": "</s>",
|
16 |
+
"lstrip": false,
|
17 |
+
"normalized": false,
|
18 |
+
"rstrip": false,
|
19 |
+
"single_word": false
|
20 |
+
},
|
21 |
+
"legacy": false,
|
22 |
+
"model_max_length": 4096,
|
23 |
+
"pad_token": null,
|
24 |
+
"padding_side": "right",
|
25 |
+
"sp_model_kwargs": {},
|
26 |
+
"spaces_between_special_tokens": false,
|
27 |
+
"tokenizer_class": "LlamaTokenizer",
|
28 |
+
"unk_token": {
|
29 |
+
"__type": "AddedToken",
|
30 |
+
"content": "<unk>",
|
31 |
+
"lstrip": false,
|
32 |
+
"normalized": false,
|
33 |
+
"rstrip": false,
|
34 |
+
"single_word": false
|
35 |
+
},
|
36 |
+
"use_default_system_prompt": true
|
37 |
+
}
|