LoneStriker commited on
Commit
e760cf9
1 Parent(s): ff5be5c

Upload folder using huggingface_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
+ bagel.png filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,404 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ datasets:
4
+ - ai2_arc
5
+ - jondurbin/airoboros-3.2
6
+ - codeparrot/apps
7
+ - facebook/belebele
8
+ - boolq
9
+ - jondurbin/cinematika-v0.1
10
+ - drop
11
+ - lmsys/lmsys-chat-1m
12
+ - TIGER-Lab/MathInstruct
13
+ - cais/mmlu
14
+ - Muennighoff/natural-instructions
15
+ - openbookqa
16
+ - piqa
17
+ - Vezora/Tested-22k-Python-Alpaca
18
+ - cakiki/rosetta-code
19
+ - Open-Orca/SlimOrca
20
+ - spider
21
+ - squad_v2
22
+ - migtissera/Synthia-v1.3
23
+ - datasets/winogrande
24
+ - nvidia/HelpSteer
25
+ - Intel/orca_dpo_pairs
26
+ - unalignment/toxic-dpo-v0.1
27
+ - jondurbin/truthy-dpo-v0.1
28
+ - allenai/ultrafeedback_binarized_cleaned
29
+ - Squish42/bluemoon-fandom-1-1-rp-cleaned
30
+ - LDJnr/Capybara
31
+ - JULIELab/EmoBank
32
+ - kingbri/PIPPA-shareGPT
33
+ ---
34
+
35
+ # A bagel, with everything (except DPO)
36
+
37
+ ![bagel](bagel.png)
38
+
39
+ ## Overview
40
+
41
+ An experimental fine-tune of [mixtral-8x7b-v0.1](https://huggingface.co/mistralai/Mixtral-8x7B-v0.1) using [bagel](https://github.com/jondurbin/bagel)
42
+
43
+ This is the model after the SFT phase, before DPO has been applied.
44
+
45
+ Hardware kindly provided by [Massed Compute](https://massedcompute.com/?utm_source=huggingface&utm_creative_format=model_card&utm_content=creator_jon)
46
+
47
+ ### Data sources
48
+
49
+ *Yes, you will see benchmark names in the list, but this only uses the train splits, and a decontamination by cosine similarity is performed at the end as a sanity check*
50
+
51
+ - [ai2_arc](https://huggingface.co/datasets/ai2_arc)
52
+ - Abstraction and reasoning dataset, useful in measuring "intelligence" to a certain extent.
53
+ - [airoboros](https://huggingface.co/datasets/unalignment/spicy-3.1)
54
+ - Variety of categories of synthetic instructions generated by gpt-4.
55
+ - [apps](https://huggingface.co/datasets/codeparrot/apps)
56
+ - Python coding dataset with 10k problems.
57
+ - [belebele](https://huggingface.co/datasets/facebook/belebele)
58
+ - Multi-lingual reading comprehension dataset.
59
+ - [bluemoon](https://huggingface.co/datasets/Squish42/bluemoon-fandom-1-1-rp-cleaned)
60
+ - Roleplay data scraped from Bluemoon, then cleaned and formatted as ShareGPT.
61
+ - [boolq](https://huggingface.co/datasets/boolq)
62
+ - Corpus of yes/no questions (which can be surprisingly difficult for AI to answer apparently?)
63
+ - [capybara](https://huggingface.co/datasets/LDJnr/Capybara)
64
+ - Multi-turn dataset used to create the capybara models.
65
+ - [cinematika](https://huggingface.co/datasets/jondurbin/cinematika-v0.1) (instruction and plain text)
66
+ - RP-style data synthesized from movie scripts so the model isn't quite as boring as it otherwise would be.
67
+ - [drop](https://huggingface.co/datasets/drop)
68
+ - More reading comprehension.
69
+ - [emobank](https://github.com/JULIELab/EmoBank)
70
+ - Emotion annotations using the Valence-Arousal-Domninance scheme.
71
+ - [gutenberg](https://www.gutenberg.org/) (plain text)
72
+ - Books/plain text, again to make the model less boring, only a handful of examples supported by [chapterize](https://github.com/JonathanReeve/chapterize)
73
+ - [lmsys_chat_1m](https://huggingface.co/datasets/lmsys/lmsys-chat-1m) (only gpt-4 items, also used for DPO)
74
+ - Chats collected by the lmsys chat arena, containing a wide variety of chats with various models.
75
+ - [mathinstruct](https://huggingface.co/datasets/TIGER-Lab/MathInstruct)
76
+ - Composite dataset with a variety of math-related tasks and problem/question formats.
77
+ - [mmlu](https://huggingface.co/datasets/cais/mmlu)
78
+ - Massive Multitask Language Understanding - a wide variety of questions about various subject matters.
79
+ - [natural_instructions](https://huggingface.co/datasets/Muennighoff/natural-instructions)
80
+ - Millions of instructions from 1600+ task categories (sampled down substantially, stratified by task type)
81
+ - [openbookqa](https://huggingface.co/datasets/openbookqa)
82
+ - Question answering dataset.
83
+ - [pippa](https://huggingface.co/datasets/kingbri/PIPPA-shareGPT)
84
+ - Deduped version of [PIPPA](https://huggingface.co/datasets/PygmalionAI/PIPPA) in ShareGPT format.
85
+ - [piqa](https://huggingface.co/datasets/piqa)
86
+ - Phyiscal interaction question answering.
87
+ - [python_alpaca](https://huggingface.co/datasets/Vezora/Tested-22k-Python-Alpaca)
88
+ - Python instruction response pairs, validated as functional.
89
+ - [rosetta_code](https://huggingface.co/datasets/cakiki/rosetta-code)
90
+ - Code problems and solutions in a variety of programming languages taken from rosettacode.org.
91
+ - [slimorca](https://huggingface.co/datasets/Open-Orca/SlimOrca)
92
+ - Collection of ~500k gpt-4 verified chats from OpenOrca.
93
+ - [spider](https://huggingface.co/datasets/spider)
94
+ - SQL-targeted dataset.
95
+ - [squad_v2](https://huggingface.co/datasets/squad_v2)
96
+ - Contextual question answering (RAG).
97
+ - [synthia](https://huggingface.co/datasets/migtissera/Synthia-v1.3)
98
+ - GPT-4 generated data using advanced prompting from Migel Tissera.
99
+ - [winogrande](https://huggingface.co/datasets/winogrande)
100
+ - Fill in the blank style prompts.
101
+
102
+ Only the train splits were used (if a split was provided), and an additional pass of decontamination is performed using approximate nearest neighbor search (via faiss).
103
+
104
+ ## Prompt formatting
105
+
106
+ In sticking with the theme of the bagel, I didn't want to use a single prompt format, so I used 4 - vicuna, llama-2, alpaca, and chat-ml (sorta).
107
+ I also didn't want to randomly select a single prompt format for each item (hoping each instruction would generalize more when used in a variety of prompt formats), so each instruction is actually converted into every prompt format.
108
+
109
+ This means each epoch of our fine-tune is really basically 4 epochs. So, for the fine-tunes, I would recommend only doing 1 epoch (or 0.75 epochs). I am testing with a single epoch using a relatively low learning rate.
110
+
111
+ ### Alpaca (sort of)
112
+
113
+ ```
114
+ Below is an instruction that describes a task. Write a response that appropriately completes the request.
115
+
116
+ ### Instruction:
117
+ {system prompt, if provided}
118
+ {instruction}
119
+
120
+ ### Response:
121
+ ```
122
+
123
+ The main difference here is that because of the dataset formatting and variety of data sources, it would have been much to tedious to add an `### Input:` block, so the inputs are just in the instruction section.
124
+
125
+ ### Vicuna
126
+
127
+ ```
128
+ {system prompt, if provided, randomly defaulting to "A chat between a user and an unbiased, uncensored assistant."}
129
+ USER: {instruction}
130
+ ASSISTANT:
131
+ ```
132
+
133
+ ### ChatML (sort of)
134
+
135
+ I don't really understand the point of having special tokens for `<|im_start|>` and `<|im_end|>`, because in practice they just act as BOS and EOS tokens (but, please correct me if I'm wrong).
136
+
137
+ So, instead of:
138
+ ```text
139
+ {bos}<|im_start|>{role}
140
+ {text}
141
+ <|im_end|>{eos}
142
+ ```
143
+
144
+ I just changed it to:
145
+ ```text
146
+ {bos}{role}
147
+ {text}
148
+ {eos}
149
+ ```
150
+
151
+ If you *really* want to use `<|im_start|>` and `<|im_end|>`, just update your `tokenizer_config.json` to use `<|im_start|>` instead of `<s>` and `<|im_end|>` instead of `</s>` and when tokenizing. And if you still don't like what I've done to this chat-ml-ish format, feel free to cry into your pillow or fork the code and do a new fine-tune.
152
+
153
+ ### Llama-2 chat
154
+
155
+ ```
156
+ [INST] <<SYS>>
157
+ {system}
158
+ <</SYS>>
159
+
160
+ {instruction} [/INST]
161
+ ```
162
+
163
+ ### Default via chat template
164
+
165
+ The model's `tokenizer_config.json` includes the default chat template (llama-2), so you can simply use the `apply_chat_template` method to build the full prompt.
166
+
167
+ ```
168
+ import transformers
169
+ tokenizer = transformers.AutoTokenizer.from_pretrained('jondurbin/bagel-8x7b-v0.2')
170
+ chat = [
171
+ {"role": "system", "content": "You are Bob, a friendly AI assistant."},
172
+ {"role": "user", "content": "Hello, how are you?"},
173
+ {"role": "assistant", "content": "I'm doing great. How can I help you today?"},
174
+ {"role": "user", "content": "I'd like to show off how chat templating works!"},
175
+ ]
176
+ print(tokenizer.apply_chat_template(chat, tokenize=False))
177
+ ```
178
+
179
+ ### Contribute
180
+
181
+ If you're interested in new functionality/datasets, take a look at [bagel repo](https://github.com/jondurbin/bagel) and either make a PR or open an issue with details.
182
+
183
+ To help me with the fine-tuning costs (which are extremely expensive for these large combined datasets):
184
+
185
+ - https://bmc.link/jondurbin
186
+ - ETH 0xce914eAFC2fe52FdceE59565Dd92c06f776fcb11
187
+ - BTC bc1qdwuth4vlg8x37ggntlxu5cjfwgmdy5zaa7pswf
188
+
189
+ ### Guide for certain tasks
190
+
191
+ #### RA(G)/contextual question answering
192
+
193
+ The model was trained to ignore what it thinks it knows, and uses the context to answer the questions, when using the format below.
194
+ The model was also tuned to limit the values to the provided context as much as possible to reduce hallucinations.
195
+
196
+ The format for a contextual prompt is as follows:
197
+ ```
198
+ BEGININPUT
199
+ BEGINCONTEXT
200
+ [key0: value0]
201
+ [key1: value1]
202
+ ... other metdata ...
203
+ ENDCONTEXT
204
+ [insert your text blocks here]
205
+ ENDINPUT
206
+ [add as many other blocks, in the exact same format]
207
+ BEGININSTRUCTION
208
+ [insert your instruction(s). The model was tuned with single questions, paragraph format, lists, etc.]
209
+ ENDINSTRUCTION
210
+ ```
211
+
212
+ I know it's a bit verbose and annoying, but after much trial and error, using these explicit delimiters helps the model understand where to find the responses and how to associate specific sources with it.
213
+ - `BEGININPUT` - denotes a new input block
214
+ - `BEGINCONTEXT` - denotes the block of context (metadata key/value pairs) to associate with the current input block
215
+ - `ENDCONTEXT` - denotes the end of the metadata block for the current input
216
+ - [text] - Insert whatever text you want for the input block, as many paragraphs as can fit in the context.
217
+ - `ENDINPUT` - denotes the end of the current input block
218
+ - [repeat as many input blocks in this format as you want]
219
+ - `BEGININSTRUCTION` - denotes the start of the list (or one) instruction(s) to respond to for all of the input blocks above.
220
+ - [instruction(s)]
221
+ - `ENDINSTRUCTION` - denotes the end of instruction set
222
+
223
+ __Use a very low temperature!__
224
+
225
+ Here's a trivial, but important example to prove the point:
226
+ ```
227
+ BEGININPUT
228
+ BEGINCONTEXT
229
+ date: 2021-01-01
230
+ url: https://web.site/123
231
+ ENDCONTEXT
232
+ In a shocking turn of events, blueberries are now green, but will be sticking with the same name.
233
+ ENDINPUT
234
+ BEGININSTRUCTION
235
+ What color are bluberries? Source?
236
+ ENDINSTRUCTION
237
+ ```
238
+
239
+ And the response:
240
+ ```
241
+ Blueberries are now green.
242
+ Source:
243
+ date: 2021-01-01
244
+ url: https://web.site/123
245
+ ```
246
+
247
+ #### Summarization
248
+
249
+ 500 samples have been included from [this dataset](https://huggingface.co/datasets/mattpscott/airoboros-summarization), using the same format as contextual question answering, for example:
250
+
251
+ ```
252
+ BEGININPUT
253
+ {text to summarize}
254
+ ENDINPUT
255
+ BEGININSTRUCTION
256
+ Summarize the input in around 130 words.
257
+ ENDINSTRUCTION
258
+ ```
259
+
260
+ #### Agent/function calling
261
+
262
+ The dataset includes many examples of function/args generation based on input criteria. This is somewhat similar to the OpenAI function calling, but the output is either JSON or YAML.
263
+
264
+ Example prompt:
265
+ ```
266
+ As an AI assistant, please select the most suitable function and parameters from the list of available functions below, based on the user's input. Provide your response in JSON format.
267
+
268
+ Input: I want to know how many times 'Python' is mentioned in my text file.
269
+
270
+ Available functions:
271
+ file_analytics:
272
+ description: This tool performs various operations on a text file.
273
+ params:
274
+ action: The operation we want to perform on the data, such as "count_occurrences", "find_line", etc.
275
+ filters:
276
+ keyword: The word or phrase we want to search for.
277
+ ```
278
+
279
+ Response:
280
+ ```json
281
+ {
282
+ "function": "file_analytics",
283
+ "params": {
284
+ "action": "count_occurrences",
285
+ "filters": {
286
+ "keyword": "Python"
287
+ }
288
+ }
289
+ }
290
+ ```
291
+
292
+ #### reWOO style execution planning
293
+
294
+ The model now supports execution planning for complex instructions that would require making use of several tools. The output is just the plan, you must implement a mechanism to parse the output and actually call the functions!
295
+
296
+ Example prompt:
297
+ ```
298
+ Please construct a systematic plan to generate an optimal response to the user instruction, utilizing a set of provided tools. Each plan will correspond to an evidence value, which will be the output of one of the available functions given an input string
299
+ that could be the user's question, one or more prior evidence values, or a combination of both.
300
+
301
+ Here are the tools available to be called:
302
+ DuckDuckGo[input]: Tool that enables the user to search for information using the DuckDuckGo search engine. This tool is helpful in locating a suitable list of sites that may contain the answer to the user's query. It does not directly answer the question but finds a list of sites that might have the answer.
303
+ TextScraper[input]: Load one or more websites from the input string containing newline delimited links, where input is one or more links, and produces plain text output containing the content of the links.
304
+ HyperlinkExtractor[input]: Extract links from plain text and produces a plain text, newline delimited response of links.
305
+ KnowledgeModel[input]: Question answering language model, particularly useful in answering questions based on an input passage of text. The input must be a text question that references an :evidence[n]: variable, e.g. What color is the cat, given :evidence1:?
306
+
307
+ The output should be in format:
308
+ Plan: [first action to take based on input question]
309
+ :evidence0: = [function to call with input parameter]
310
+ Plan: [next action to take, based on result of :evidence0:]
311
+ :evidence1: = [next function to call and input parameter, which may include reference to previous evidence, e.g. "Given context :evidence0"]
312
+ ...
313
+ Answer: [:evidence[n]: containing the final answer.]
314
+
315
+ Question: Who won the Nobel Prize in Literature in 2020 and what is their most famous work?
316
+ ```
317
+
318
+ Response:
319
+ ```
320
+ Plan: Begin by conducting a web search to find out who won the Nobel Prize in Literature in 2020.
321
+ :evidence0: = DuckDuckGo[Nobel Prize in Literature 2020 winner]
322
+ Plan: Utilize the KnowledgeModel to interpret the search results and find the name of the laureate.
323
+ :evidence1: = KnowledgeModel[Who won the Nobel Prize in Literature in 2020, given :evidence0:?]
324
+ Plan: Conduct another web search to find the most famous work of the identified laureate.
325
+ :evidence2: = DuckDuckGo[Most famous work of :evidence1:]
326
+ Plan: Extract the relevant links from the DuckDuckGo search results for a more focused search.
327
+ :evidence3: = HyperlinkExtractor[:evidence2:]
328
+ Plan: Use the TextScraper tool to extract information from the relevant links.
329
+ :evidence4: = TextScraper[:evidence3:]
330
+ Plan: Finally, utilize the KnowledgeModel to identify and summarize the most famous work of the laureate from the extracted information.
331
+ :evidence5: = KnowledgeModel[What is the most famous work of :evidence1:, given :evidence4:?]
332
+ Answer: :evidence5:
333
+ ```
334
+
335
+ For this to be useful, you'd have to parse the output plan text, and implement/call each of the functions. This is just pseudo-code, completely untested off the top of my head, and obviously would requiring full implementation + hardening:
336
+
337
+ ```python
338
+ import re
339
+ import requests
340
+
341
+ def inject_context(input_text, **context):
342
+ for ref in set(re.findall(r"(:evidence[0-9]+:)", input_text, re.I)):
343
+ input_text = input_text.replace(ref, context.get(ref, ""))
344
+ return input_text
345
+
346
+ def duckduckgo(input_text, **context):
347
+ search_string = inject_context(input_text, **context)
348
+ ... search via duck duck go using search_string
349
+ ... return text content
350
+
351
+ def link_extractor(input_text, **context):
352
+ input_text = inject_context(input_text, **context)
353
+ return "\n".join(list(set(re.findall(r"(https?://[^\s]+?\.?)", input_text, re.I))))
354
+
355
+ def scrape(input_text, **context):
356
+ input_text = inject_context(input_text, **context)
357
+ text = []
358
+ for link in input_text.splitlines():
359
+ text.append(requests.get(link).text)
360
+ return "\n".join(text)
361
+
362
+ def infer(input_text, **context)
363
+ prompt = inject_context(input_text, **context)
364
+ ... call model with prompt, return output
365
+
366
+ def parse_plan(plan):
367
+ method_map = {
368
+ "DuckDuckGo": duckduckgo,
369
+ "HyperlinkExtractor": link_extractor,
370
+ "KnowledgeModel": infer,
371
+ "TextScraper": scrape,
372
+ }
373
+ context = {}
374
+ for line in plan.strip().splitlines():
375
+ if line.startswith("Plan:"):
376
+ print(line)
377
+ continue
378
+ parts = re.match("^(:evidence[0-9]+:)\s*=\s*([^\[]+])(\[.*\])\s$", line, re.I)
379
+ if not parts:
380
+ if line.startswith("Answer: "):
381
+ return context.get(line.split(" ")[-1].strip(), "Answer couldn't be generated...")
382
+ raise RuntimeError("bad format: " + line)
383
+ context[parts.group(1)] = method_map[parts.group(2)](parts.group(3), **context)
384
+ ```
385
+
386
+ ### Fine-tuning information
387
+
388
+ You can find charts, and the full configuration used to fine-tune this model on [weights and biases](https://wandb.ai/jondurbin/bagel-8x7b-v0.2/runs/agxjjdso?workspace=user-jondurbin)
389
+
390
+ The model was fine-tuned on an 8x a6000 instance, for 4 days, 15 hours, 6 minutes and 42 seconds.
391
+
392
+ ### Licence and usage restrictions
393
+
394
+ The base model is mixtral-8x7b-v0.1, which is licensed as apache-2.0 - no issues there.
395
+
396
+ The fine-tuning data, however, includes several datasets that have data generated at least in part by OpenAI's gpt-4.
397
+
398
+ I am not a lawyer, so I can't help determine if this is actually commercially viable, but some questions that often come up are:
399
+
400
+ - Does the OpenAI ToS apply only to the user who created the dataset initially, and not subsequent models?
401
+ - If the dataset was released under a permissive license, but actually includes OpenAI generated data, does that ToS supersede the license?
402
+ - Does the dataset fall completely under fair use anyways, since the model isn't really capable of reproducing the entire training set verbatim?
403
+
404
+ Use your best judgement and seek legal advice if you are concerned about the terms. In any case, by using this model, you agree to completely indemnify me.
bagel.png ADDED

Git LFS Details

  • SHA256: 9d922a78a6f7d2de37f094d9eef558fd87dfc8e8df293c195aae27cb402b4160
  • Pointer size: 132 Bytes
  • Size of remote file: 2.15 MB
config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "bagel-8x7b-v0.2",
3
+ "architectures": [
4
+ "MixtralForCausalLM"
5
+ ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 1,
8
+ "eos_token_id": 2,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 4096,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 14336,
13
+ "max_position_embeddings": 32768,
14
+ "model_type": "mixtral",
15
+ "num_attention_heads": 32,
16
+ "num_experts_per_tok": 2,
17
+ "num_hidden_layers": 32,
18
+ "num_key_value_heads": 8,
19
+ "num_local_experts": 8,
20
+ "output_router_logits": false,
21
+ "rms_norm_eps": 1e-05,
22
+ "rope_theta": 1000000.0,
23
+ "router_aux_loss_coef": 0.02,
24
+ "sliding_window": null,
25
+ "tie_word_embeddings": false,
26
+ "torch_dtype": "bfloat16",
27
+ "transformers_version": "4.37.0.dev0",
28
+ "use_cache": true,
29
+ "vocab_size": 32000
30
+ }
generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 2,
5
+ "transformers_version": "4.37.0.dev0"
6
+ }
model.safetensors.index.json ADDED
@@ -0,0 +1,1002 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 93405585408
4
+ },
5
+ "weight_map": {
6
+ "lm_head.weight": "model-00024-of-00024.safetensors",
7
+ "model.embed_tokens.weight": "model-00001-of-00024.safetensors",
8
+ "model.layers.0.block_sparse_moe.experts.0.w1.weight": "model-00001-of-00024.safetensors",
9
+ "model.layers.0.block_sparse_moe.experts.0.w2.weight": "model-00001-of-00024.safetensors",
10
+ "model.layers.0.block_sparse_moe.experts.0.w3.weight": "model-00001-of-00024.safetensors",
11
+ "model.layers.0.block_sparse_moe.experts.1.w1.weight": "model-00001-of-00024.safetensors",
12
+ "model.layers.0.block_sparse_moe.experts.1.w2.weight": "model-00001-of-00024.safetensors",
13
+ "model.layers.0.block_sparse_moe.experts.1.w3.weight": "model-00001-of-00024.safetensors",
14
+ "model.layers.0.block_sparse_moe.experts.2.w1.weight": "model-00001-of-00024.safetensors",
15
+ "model.layers.0.block_sparse_moe.experts.2.w2.weight": "model-00001-of-00024.safetensors",
16
+ "model.layers.0.block_sparse_moe.experts.2.w3.weight": "model-00001-of-00024.safetensors",
17
+ "model.layers.0.block_sparse_moe.experts.3.w1.weight": "model-00001-of-00024.safetensors",
18
+ "model.layers.0.block_sparse_moe.experts.3.w2.weight": "model-00001-of-00024.safetensors",
19
+ "model.layers.0.block_sparse_moe.experts.3.w3.weight": "model-00001-of-00024.safetensors",
20
+ "model.layers.0.block_sparse_moe.experts.4.w1.weight": "model-00001-of-00024.safetensors",
21
+ "model.layers.0.block_sparse_moe.experts.4.w2.weight": "model-00001-of-00024.safetensors",
22
+ "model.layers.0.block_sparse_moe.experts.4.w3.weight": "model-00001-of-00024.safetensors",
23
+ "model.layers.0.block_sparse_moe.experts.5.w1.weight": "model-00001-of-00024.safetensors",
24
+ "model.layers.0.block_sparse_moe.experts.5.w2.weight": "model-00001-of-00024.safetensors",
25
+ "model.layers.0.block_sparse_moe.experts.5.w3.weight": "model-00001-of-00024.safetensors",
26
+ "model.layers.0.block_sparse_moe.experts.6.w1.weight": "model-00001-of-00024.safetensors",
27
+ "model.layers.0.block_sparse_moe.experts.6.w2.weight": "model-00001-of-00024.safetensors",
28
+ "model.layers.0.block_sparse_moe.experts.6.w3.weight": "model-00001-of-00024.safetensors",
29
+ "model.layers.0.block_sparse_moe.experts.7.w1.weight": "model-00001-of-00024.safetensors",
30
+ "model.layers.0.block_sparse_moe.experts.7.w2.weight": "model-00001-of-00024.safetensors",
31
+ "model.layers.0.block_sparse_moe.experts.7.w3.weight": "model-00001-of-00024.safetensors",
32
+ "model.layers.0.block_sparse_moe.gate.weight": "model-00001-of-00024.safetensors",
33
+ "model.layers.0.input_layernorm.weight": "model-00001-of-00024.safetensors",
34
+ "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00024.safetensors",
35
+ "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00024.safetensors",
36
+ "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00024.safetensors",
37
+ "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00024.safetensors",
38
+ "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00024.safetensors",
39
+ "model.layers.1.block_sparse_moe.experts.0.w1.weight": "model-00001-of-00024.safetensors",
40
+ "model.layers.1.block_sparse_moe.experts.0.w2.weight": "model-00001-of-00024.safetensors",
41
+ "model.layers.1.block_sparse_moe.experts.0.w3.weight": "model-00001-of-00024.safetensors",
42
+ "model.layers.1.block_sparse_moe.experts.1.w1.weight": "model-00001-of-00024.safetensors",
43
+ "model.layers.1.block_sparse_moe.experts.1.w2.weight": "model-00001-of-00024.safetensors",
44
+ "model.layers.1.block_sparse_moe.experts.1.w3.weight": "model-00001-of-00024.safetensors",
45
+ "model.layers.1.block_sparse_moe.experts.2.w1.weight": "model-00002-of-00024.safetensors",
46
+ "model.layers.1.block_sparse_moe.experts.2.w2.weight": "model-00002-of-00024.safetensors",
47
+ "model.layers.1.block_sparse_moe.experts.2.w3.weight": "model-00002-of-00024.safetensors",
48
+ "model.layers.1.block_sparse_moe.experts.3.w1.weight": "model-00002-of-00024.safetensors",
49
+ "model.layers.1.block_sparse_moe.experts.3.w2.weight": "model-00002-of-00024.safetensors",
50
+ "model.layers.1.block_sparse_moe.experts.3.w3.weight": "model-00002-of-00024.safetensors",
51
+ "model.layers.1.block_sparse_moe.experts.4.w1.weight": "model-00002-of-00024.safetensors",
52
+ "model.layers.1.block_sparse_moe.experts.4.w2.weight": "model-00002-of-00024.safetensors",
53
+ "model.layers.1.block_sparse_moe.experts.4.w3.weight": "model-00002-of-00024.safetensors",
54
+ "model.layers.1.block_sparse_moe.experts.5.w1.weight": "model-00002-of-00024.safetensors",
55
+ "model.layers.1.block_sparse_moe.experts.5.w2.weight": "model-00002-of-00024.safetensors",
56
+ "model.layers.1.block_sparse_moe.experts.5.w3.weight": "model-00002-of-00024.safetensors",
57
+ "model.layers.1.block_sparse_moe.experts.6.w1.weight": "model-00002-of-00024.safetensors",
58
+ "model.layers.1.block_sparse_moe.experts.6.w2.weight": "model-00002-of-00024.safetensors",
59
+ "model.layers.1.block_sparse_moe.experts.6.w3.weight": "model-00002-of-00024.safetensors",
60
+ "model.layers.1.block_sparse_moe.experts.7.w1.weight": "model-00002-of-00024.safetensors",
61
+ "model.layers.1.block_sparse_moe.experts.7.w2.weight": "model-00002-of-00024.safetensors",
62
+ "model.layers.1.block_sparse_moe.experts.7.w3.weight": "model-00002-of-00024.safetensors",
63
+ "model.layers.1.block_sparse_moe.gate.weight": "model-00001-of-00024.safetensors",
64
+ "model.layers.1.input_layernorm.weight": "model-00002-of-00024.safetensors",
65
+ "model.layers.1.post_attention_layernorm.weight": "model-00002-of-00024.safetensors",
66
+ "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00024.safetensors",
67
+ "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00024.safetensors",
68
+ "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00024.safetensors",
69
+ "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00024.safetensors",
70
+ "model.layers.10.block_sparse_moe.experts.0.w1.weight": "model-00008-of-00024.safetensors",
71
+ "model.layers.10.block_sparse_moe.experts.0.w2.weight": "model-00008-of-00024.safetensors",
72
+ "model.layers.10.block_sparse_moe.experts.0.w3.weight": "model-00008-of-00024.safetensors",
73
+ "model.layers.10.block_sparse_moe.experts.1.w1.weight": "model-00008-of-00024.safetensors",
74
+ "model.layers.10.block_sparse_moe.experts.1.w2.weight": "model-00008-of-00024.safetensors",
75
+ "model.layers.10.block_sparse_moe.experts.1.w3.weight": "model-00008-of-00024.safetensors",
76
+ "model.layers.10.block_sparse_moe.experts.2.w1.weight": "model-00008-of-00024.safetensors",
77
+ "model.layers.10.block_sparse_moe.experts.2.w2.weight": "model-00008-of-00024.safetensors",
78
+ "model.layers.10.block_sparse_moe.experts.2.w3.weight": "model-00008-of-00024.safetensors",
79
+ "model.layers.10.block_sparse_moe.experts.3.w1.weight": "model-00008-of-00024.safetensors",
80
+ "model.layers.10.block_sparse_moe.experts.3.w2.weight": "model-00008-of-00024.safetensors",
81
+ "model.layers.10.block_sparse_moe.experts.3.w3.weight": "model-00008-of-00024.safetensors",
82
+ "model.layers.10.block_sparse_moe.experts.4.w1.weight": "model-00008-of-00024.safetensors",
83
+ "model.layers.10.block_sparse_moe.experts.4.w2.weight": "model-00008-of-00024.safetensors",
84
+ "model.layers.10.block_sparse_moe.experts.4.w3.weight": "model-00008-of-00024.safetensors",
85
+ "model.layers.10.block_sparse_moe.experts.5.w1.weight": "model-00008-of-00024.safetensors",
86
+ "model.layers.10.block_sparse_moe.experts.5.w2.weight": "model-00008-of-00024.safetensors",
87
+ "model.layers.10.block_sparse_moe.experts.5.w3.weight": "model-00008-of-00024.safetensors",
88
+ "model.layers.10.block_sparse_moe.experts.6.w1.weight": "model-00008-of-00024.safetensors",
89
+ "model.layers.10.block_sparse_moe.experts.6.w2.weight": "model-00009-of-00024.safetensors",
90
+ "model.layers.10.block_sparse_moe.experts.6.w3.weight": "model-00009-of-00024.safetensors",
91
+ "model.layers.10.block_sparse_moe.experts.7.w1.weight": "model-00009-of-00024.safetensors",
92
+ "model.layers.10.block_sparse_moe.experts.7.w2.weight": "model-00009-of-00024.safetensors",
93
+ "model.layers.10.block_sparse_moe.experts.7.w3.weight": "model-00009-of-00024.safetensors",
94
+ "model.layers.10.block_sparse_moe.gate.weight": "model-00008-of-00024.safetensors",
95
+ "model.layers.10.input_layernorm.weight": "model-00009-of-00024.safetensors",
96
+ "model.layers.10.post_attention_layernorm.weight": "model-00009-of-00024.safetensors",
97
+ "model.layers.10.self_attn.k_proj.weight": "model-00008-of-00024.safetensors",
98
+ "model.layers.10.self_attn.o_proj.weight": "model-00008-of-00024.safetensors",
99
+ "model.layers.10.self_attn.q_proj.weight": "model-00008-of-00024.safetensors",
100
+ "model.layers.10.self_attn.v_proj.weight": "model-00008-of-00024.safetensors",
101
+ "model.layers.11.block_sparse_moe.experts.0.w1.weight": "model-00009-of-00024.safetensors",
102
+ "model.layers.11.block_sparse_moe.experts.0.w2.weight": "model-00009-of-00024.safetensors",
103
+ "model.layers.11.block_sparse_moe.experts.0.w3.weight": "model-00009-of-00024.safetensors",
104
+ "model.layers.11.block_sparse_moe.experts.1.w1.weight": "model-00009-of-00024.safetensors",
105
+ "model.layers.11.block_sparse_moe.experts.1.w2.weight": "model-00009-of-00024.safetensors",
106
+ "model.layers.11.block_sparse_moe.experts.1.w3.weight": "model-00009-of-00024.safetensors",
107
+ "model.layers.11.block_sparse_moe.experts.2.w1.weight": "model-00009-of-00024.safetensors",
108
+ "model.layers.11.block_sparse_moe.experts.2.w2.weight": "model-00009-of-00024.safetensors",
109
+ "model.layers.11.block_sparse_moe.experts.2.w3.weight": "model-00009-of-00024.safetensors",
110
+ "model.layers.11.block_sparse_moe.experts.3.w1.weight": "model-00009-of-00024.safetensors",
111
+ "model.layers.11.block_sparse_moe.experts.3.w2.weight": "model-00009-of-00024.safetensors",
112
+ "model.layers.11.block_sparse_moe.experts.3.w3.weight": "model-00009-of-00024.safetensors",
113
+ "model.layers.11.block_sparse_moe.experts.4.w1.weight": "model-00009-of-00024.safetensors",
114
+ "model.layers.11.block_sparse_moe.experts.4.w2.weight": "model-00009-of-00024.safetensors",
115
+ "model.layers.11.block_sparse_moe.experts.4.w3.weight": "model-00009-of-00024.safetensors",
116
+ "model.layers.11.block_sparse_moe.experts.5.w1.weight": "model-00009-of-00024.safetensors",
117
+ "model.layers.11.block_sparse_moe.experts.5.w2.weight": "model-00009-of-00024.safetensors",
118
+ "model.layers.11.block_sparse_moe.experts.5.w3.weight": "model-00009-of-00024.safetensors",
119
+ "model.layers.11.block_sparse_moe.experts.6.w1.weight": "model-00009-of-00024.safetensors",
120
+ "model.layers.11.block_sparse_moe.experts.6.w2.weight": "model-00009-of-00024.safetensors",
121
+ "model.layers.11.block_sparse_moe.experts.6.w3.weight": "model-00009-of-00024.safetensors",
122
+ "model.layers.11.block_sparse_moe.experts.7.w1.weight": "model-00009-of-00024.safetensors",
123
+ "model.layers.11.block_sparse_moe.experts.7.w2.weight": "model-00009-of-00024.safetensors",
124
+ "model.layers.11.block_sparse_moe.experts.7.w3.weight": "model-00009-of-00024.safetensors",
125
+ "model.layers.11.block_sparse_moe.gate.weight": "model-00009-of-00024.safetensors",
126
+ "model.layers.11.input_layernorm.weight": "model-00009-of-00024.safetensors",
127
+ "model.layers.11.post_attention_layernorm.weight": "model-00009-of-00024.safetensors",
128
+ "model.layers.11.self_attn.k_proj.weight": "model-00009-of-00024.safetensors",
129
+ "model.layers.11.self_attn.o_proj.weight": "model-00009-of-00024.safetensors",
130
+ "model.layers.11.self_attn.q_proj.weight": "model-00009-of-00024.safetensors",
131
+ "model.layers.11.self_attn.v_proj.weight": "model-00009-of-00024.safetensors",
132
+ "model.layers.12.block_sparse_moe.experts.0.w1.weight": "model-00009-of-00024.safetensors",
133
+ "model.layers.12.block_sparse_moe.experts.0.w2.weight": "model-00009-of-00024.safetensors",
134
+ "model.layers.12.block_sparse_moe.experts.0.w3.weight": "model-00009-of-00024.safetensors",
135
+ "model.layers.12.block_sparse_moe.experts.1.w1.weight": "model-00010-of-00024.safetensors",
136
+ "model.layers.12.block_sparse_moe.experts.1.w2.weight": "model-00010-of-00024.safetensors",
137
+ "model.layers.12.block_sparse_moe.experts.1.w3.weight": "model-00010-of-00024.safetensors",
138
+ "model.layers.12.block_sparse_moe.experts.2.w1.weight": "model-00010-of-00024.safetensors",
139
+ "model.layers.12.block_sparse_moe.experts.2.w2.weight": "model-00010-of-00024.safetensors",
140
+ "model.layers.12.block_sparse_moe.experts.2.w3.weight": "model-00010-of-00024.safetensors",
141
+ "model.layers.12.block_sparse_moe.experts.3.w1.weight": "model-00010-of-00024.safetensors",
142
+ "model.layers.12.block_sparse_moe.experts.3.w2.weight": "model-00010-of-00024.safetensors",
143
+ "model.layers.12.block_sparse_moe.experts.3.w3.weight": "model-00010-of-00024.safetensors",
144
+ "model.layers.12.block_sparse_moe.experts.4.w1.weight": "model-00010-of-00024.safetensors",
145
+ "model.layers.12.block_sparse_moe.experts.4.w2.weight": "model-00010-of-00024.safetensors",
146
+ "model.layers.12.block_sparse_moe.experts.4.w3.weight": "model-00010-of-00024.safetensors",
147
+ "model.layers.12.block_sparse_moe.experts.5.w1.weight": "model-00010-of-00024.safetensors",
148
+ "model.layers.12.block_sparse_moe.experts.5.w2.weight": "model-00010-of-00024.safetensors",
149
+ "model.layers.12.block_sparse_moe.experts.5.w3.weight": "model-00010-of-00024.safetensors",
150
+ "model.layers.12.block_sparse_moe.experts.6.w1.weight": "model-00010-of-00024.safetensors",
151
+ "model.layers.12.block_sparse_moe.experts.6.w2.weight": "model-00010-of-00024.safetensors",
152
+ "model.layers.12.block_sparse_moe.experts.6.w3.weight": "model-00010-of-00024.safetensors",
153
+ "model.layers.12.block_sparse_moe.experts.7.w1.weight": "model-00010-of-00024.safetensors",
154
+ "model.layers.12.block_sparse_moe.experts.7.w2.weight": "model-00010-of-00024.safetensors",
155
+ "model.layers.12.block_sparse_moe.experts.7.w3.weight": "model-00010-of-00024.safetensors",
156
+ "model.layers.12.block_sparse_moe.gate.weight": "model-00009-of-00024.safetensors",
157
+ "model.layers.12.input_layernorm.weight": "model-00010-of-00024.safetensors",
158
+ "model.layers.12.post_attention_layernorm.weight": "model-00010-of-00024.safetensors",
159
+ "model.layers.12.self_attn.k_proj.weight": "model-00009-of-00024.safetensors",
160
+ "model.layers.12.self_attn.o_proj.weight": "model-00009-of-00024.safetensors",
161
+ "model.layers.12.self_attn.q_proj.weight": "model-00009-of-00024.safetensors",
162
+ "model.layers.12.self_attn.v_proj.weight": "model-00009-of-00024.safetensors",
163
+ "model.layers.13.block_sparse_moe.experts.0.w1.weight": "model-00010-of-00024.safetensors",
164
+ "model.layers.13.block_sparse_moe.experts.0.w2.weight": "model-00010-of-00024.safetensors",
165
+ "model.layers.13.block_sparse_moe.experts.0.w3.weight": "model-00010-of-00024.safetensors",
166
+ "model.layers.13.block_sparse_moe.experts.1.w1.weight": "model-00010-of-00024.safetensors",
167
+ "model.layers.13.block_sparse_moe.experts.1.w2.weight": "model-00010-of-00024.safetensors",
168
+ "model.layers.13.block_sparse_moe.experts.1.w3.weight": "model-00010-of-00024.safetensors",
169
+ "model.layers.13.block_sparse_moe.experts.2.w1.weight": "model-00010-of-00024.safetensors",
170
+ "model.layers.13.block_sparse_moe.experts.2.w2.weight": "model-00010-of-00024.safetensors",
171
+ "model.layers.13.block_sparse_moe.experts.2.w3.weight": "model-00010-of-00024.safetensors",
172
+ "model.layers.13.block_sparse_moe.experts.3.w1.weight": "model-00010-of-00024.safetensors",
173
+ "model.layers.13.block_sparse_moe.experts.3.w2.weight": "model-00010-of-00024.safetensors",
174
+ "model.layers.13.block_sparse_moe.experts.3.w3.weight": "model-00010-of-00024.safetensors",
175
+ "model.layers.13.block_sparse_moe.experts.4.w1.weight": "model-00011-of-00024.safetensors",
176
+ "model.layers.13.block_sparse_moe.experts.4.w2.weight": "model-00011-of-00024.safetensors",
177
+ "model.layers.13.block_sparse_moe.experts.4.w3.weight": "model-00011-of-00024.safetensors",
178
+ "model.layers.13.block_sparse_moe.experts.5.w1.weight": "model-00011-of-00024.safetensors",
179
+ "model.layers.13.block_sparse_moe.experts.5.w2.weight": "model-00011-of-00024.safetensors",
180
+ "model.layers.13.block_sparse_moe.experts.5.w3.weight": "model-00011-of-00024.safetensors",
181
+ "model.layers.13.block_sparse_moe.experts.6.w1.weight": "model-00011-of-00024.safetensors",
182
+ "model.layers.13.block_sparse_moe.experts.6.w2.weight": "model-00011-of-00024.safetensors",
183
+ "model.layers.13.block_sparse_moe.experts.6.w3.weight": "model-00011-of-00024.safetensors",
184
+ "model.layers.13.block_sparse_moe.experts.7.w1.weight": "model-00011-of-00024.safetensors",
185
+ "model.layers.13.block_sparse_moe.experts.7.w2.weight": "model-00011-of-00024.safetensors",
186
+ "model.layers.13.block_sparse_moe.experts.7.w3.weight": "model-00011-of-00024.safetensors",
187
+ "model.layers.13.block_sparse_moe.gate.weight": "model-00010-of-00024.safetensors",
188
+ "model.layers.13.input_layernorm.weight": "model-00011-of-00024.safetensors",
189
+ "model.layers.13.post_attention_layernorm.weight": "model-00011-of-00024.safetensors",
190
+ "model.layers.13.self_attn.k_proj.weight": "model-00010-of-00024.safetensors",
191
+ "model.layers.13.self_attn.o_proj.weight": "model-00010-of-00024.safetensors",
192
+ "model.layers.13.self_attn.q_proj.weight": "model-00010-of-00024.safetensors",
193
+ "model.layers.13.self_attn.v_proj.weight": "model-00010-of-00024.safetensors",
194
+ "model.layers.14.block_sparse_moe.experts.0.w1.weight": "model-00011-of-00024.safetensors",
195
+ "model.layers.14.block_sparse_moe.experts.0.w2.weight": "model-00011-of-00024.safetensors",
196
+ "model.layers.14.block_sparse_moe.experts.0.w3.weight": "model-00011-of-00024.safetensors",
197
+ "model.layers.14.block_sparse_moe.experts.1.w1.weight": "model-00011-of-00024.safetensors",
198
+ "model.layers.14.block_sparse_moe.experts.1.w2.weight": "model-00011-of-00024.safetensors",
199
+ "model.layers.14.block_sparse_moe.experts.1.w3.weight": "model-00011-of-00024.safetensors",
200
+ "model.layers.14.block_sparse_moe.experts.2.w1.weight": "model-00011-of-00024.safetensors",
201
+ "model.layers.14.block_sparse_moe.experts.2.w2.weight": "model-00011-of-00024.safetensors",
202
+ "model.layers.14.block_sparse_moe.experts.2.w3.weight": "model-00011-of-00024.safetensors",
203
+ "model.layers.14.block_sparse_moe.experts.3.w1.weight": "model-00011-of-00024.safetensors",
204
+ "model.layers.14.block_sparse_moe.experts.3.w2.weight": "model-00011-of-00024.safetensors",
205
+ "model.layers.14.block_sparse_moe.experts.3.w3.weight": "model-00011-of-00024.safetensors",
206
+ "model.layers.14.block_sparse_moe.experts.4.w1.weight": "model-00011-of-00024.safetensors",
207
+ "model.layers.14.block_sparse_moe.experts.4.w2.weight": "model-00011-of-00024.safetensors",
208
+ "model.layers.14.block_sparse_moe.experts.4.w3.weight": "model-00011-of-00024.safetensors",
209
+ "model.layers.14.block_sparse_moe.experts.5.w1.weight": "model-00011-of-00024.safetensors",
210
+ "model.layers.14.block_sparse_moe.experts.5.w2.weight": "model-00011-of-00024.safetensors",
211
+ "model.layers.14.block_sparse_moe.experts.5.w3.weight": "model-00011-of-00024.safetensors",
212
+ "model.layers.14.block_sparse_moe.experts.6.w1.weight": "model-00011-of-00024.safetensors",
213
+ "model.layers.14.block_sparse_moe.experts.6.w2.weight": "model-00011-of-00024.safetensors",
214
+ "model.layers.14.block_sparse_moe.experts.6.w3.weight": "model-00011-of-00024.safetensors",
215
+ "model.layers.14.block_sparse_moe.experts.7.w1.weight": "model-00012-of-00024.safetensors",
216
+ "model.layers.14.block_sparse_moe.experts.7.w2.weight": "model-00012-of-00024.safetensors",
217
+ "model.layers.14.block_sparse_moe.experts.7.w3.weight": "model-00012-of-00024.safetensors",
218
+ "model.layers.14.block_sparse_moe.gate.weight": "model-00011-of-00024.safetensors",
219
+ "model.layers.14.input_layernorm.weight": "model-00012-of-00024.safetensors",
220
+ "model.layers.14.post_attention_layernorm.weight": "model-00012-of-00024.safetensors",
221
+ "model.layers.14.self_attn.k_proj.weight": "model-00011-of-00024.safetensors",
222
+ "model.layers.14.self_attn.o_proj.weight": "model-00011-of-00024.safetensors",
223
+ "model.layers.14.self_attn.q_proj.weight": "model-00011-of-00024.safetensors",
224
+ "model.layers.14.self_attn.v_proj.weight": "model-00011-of-00024.safetensors",
225
+ "model.layers.15.block_sparse_moe.experts.0.w1.weight": "model-00012-of-00024.safetensors",
226
+ "model.layers.15.block_sparse_moe.experts.0.w2.weight": "model-00012-of-00024.safetensors",
227
+ "model.layers.15.block_sparse_moe.experts.0.w3.weight": "model-00012-of-00024.safetensors",
228
+ "model.layers.15.block_sparse_moe.experts.1.w1.weight": "model-00012-of-00024.safetensors",
229
+ "model.layers.15.block_sparse_moe.experts.1.w2.weight": "model-00012-of-00024.safetensors",
230
+ "model.layers.15.block_sparse_moe.experts.1.w3.weight": "model-00012-of-00024.safetensors",
231
+ "model.layers.15.block_sparse_moe.experts.2.w1.weight": "model-00012-of-00024.safetensors",
232
+ "model.layers.15.block_sparse_moe.experts.2.w2.weight": "model-00012-of-00024.safetensors",
233
+ "model.layers.15.block_sparse_moe.experts.2.w3.weight": "model-00012-of-00024.safetensors",
234
+ "model.layers.15.block_sparse_moe.experts.3.w1.weight": "model-00012-of-00024.safetensors",
235
+ "model.layers.15.block_sparse_moe.experts.3.w2.weight": "model-00012-of-00024.safetensors",
236
+ "model.layers.15.block_sparse_moe.experts.3.w3.weight": "model-00012-of-00024.safetensors",
237
+ "model.layers.15.block_sparse_moe.experts.4.w1.weight": "model-00012-of-00024.safetensors",
238
+ "model.layers.15.block_sparse_moe.experts.4.w2.weight": "model-00012-of-00024.safetensors",
239
+ "model.layers.15.block_sparse_moe.experts.4.w3.weight": "model-00012-of-00024.safetensors",
240
+ "model.layers.15.block_sparse_moe.experts.5.w1.weight": "model-00012-of-00024.safetensors",
241
+ "model.layers.15.block_sparse_moe.experts.5.w2.weight": "model-00012-of-00024.safetensors",
242
+ "model.layers.15.block_sparse_moe.experts.5.w3.weight": "model-00012-of-00024.safetensors",
243
+ "model.layers.15.block_sparse_moe.experts.6.w1.weight": "model-00012-of-00024.safetensors",
244
+ "model.layers.15.block_sparse_moe.experts.6.w2.weight": "model-00012-of-00024.safetensors",
245
+ "model.layers.15.block_sparse_moe.experts.6.w3.weight": "model-00012-of-00024.safetensors",
246
+ "model.layers.15.block_sparse_moe.experts.7.w1.weight": "model-00012-of-00024.safetensors",
247
+ "model.layers.15.block_sparse_moe.experts.7.w2.weight": "model-00012-of-00024.safetensors",
248
+ "model.layers.15.block_sparse_moe.experts.7.w3.weight": "model-00012-of-00024.safetensors",
249
+ "model.layers.15.block_sparse_moe.gate.weight": "model-00012-of-00024.safetensors",
250
+ "model.layers.15.input_layernorm.weight": "model-00012-of-00024.safetensors",
251
+ "model.layers.15.post_attention_layernorm.weight": "model-00012-of-00024.safetensors",
252
+ "model.layers.15.self_attn.k_proj.weight": "model-00012-of-00024.safetensors",
253
+ "model.layers.15.self_attn.o_proj.weight": "model-00012-of-00024.safetensors",
254
+ "model.layers.15.self_attn.q_proj.weight": "model-00012-of-00024.safetensors",
255
+ "model.layers.15.self_attn.v_proj.weight": "model-00012-of-00024.safetensors",
256
+ "model.layers.16.block_sparse_moe.experts.0.w1.weight": "model-00012-of-00024.safetensors",
257
+ "model.layers.16.block_sparse_moe.experts.0.w2.weight": "model-00012-of-00024.safetensors",
258
+ "model.layers.16.block_sparse_moe.experts.0.w3.weight": "model-00012-of-00024.safetensors",
259
+ "model.layers.16.block_sparse_moe.experts.1.w1.weight": "model-00012-of-00024.safetensors",
260
+ "model.layers.16.block_sparse_moe.experts.1.w2.weight": "model-00012-of-00024.safetensors",
261
+ "model.layers.16.block_sparse_moe.experts.1.w3.weight": "model-00013-of-00024.safetensors",
262
+ "model.layers.16.block_sparse_moe.experts.2.w1.weight": "model-00013-of-00024.safetensors",
263
+ "model.layers.16.block_sparse_moe.experts.2.w2.weight": "model-00013-of-00024.safetensors",
264
+ "model.layers.16.block_sparse_moe.experts.2.w3.weight": "model-00013-of-00024.safetensors",
265
+ "model.layers.16.block_sparse_moe.experts.3.w1.weight": "model-00013-of-00024.safetensors",
266
+ "model.layers.16.block_sparse_moe.experts.3.w2.weight": "model-00013-of-00024.safetensors",
267
+ "model.layers.16.block_sparse_moe.experts.3.w3.weight": "model-00013-of-00024.safetensors",
268
+ "model.layers.16.block_sparse_moe.experts.4.w1.weight": "model-00013-of-00024.safetensors",
269
+ "model.layers.16.block_sparse_moe.experts.4.w2.weight": "model-00013-of-00024.safetensors",
270
+ "model.layers.16.block_sparse_moe.experts.4.w3.weight": "model-00013-of-00024.safetensors",
271
+ "model.layers.16.block_sparse_moe.experts.5.w1.weight": "model-00013-of-00024.safetensors",
272
+ "model.layers.16.block_sparse_moe.experts.5.w2.weight": "model-00013-of-00024.safetensors",
273
+ "model.layers.16.block_sparse_moe.experts.5.w3.weight": "model-00013-of-00024.safetensors",
274
+ "model.layers.16.block_sparse_moe.experts.6.w1.weight": "model-00013-of-00024.safetensors",
275
+ "model.layers.16.block_sparse_moe.experts.6.w2.weight": "model-00013-of-00024.safetensors",
276
+ "model.layers.16.block_sparse_moe.experts.6.w3.weight": "model-00013-of-00024.safetensors",
277
+ "model.layers.16.block_sparse_moe.experts.7.w1.weight": "model-00013-of-00024.safetensors",
278
+ "model.layers.16.block_sparse_moe.experts.7.w2.weight": "model-00013-of-00024.safetensors",
279
+ "model.layers.16.block_sparse_moe.experts.7.w3.weight": "model-00013-of-00024.safetensors",
280
+ "model.layers.16.block_sparse_moe.gate.weight": "model-00012-of-00024.safetensors",
281
+ "model.layers.16.input_layernorm.weight": "model-00013-of-00024.safetensors",
282
+ "model.layers.16.post_attention_layernorm.weight": "model-00013-of-00024.safetensors",
283
+ "model.layers.16.self_attn.k_proj.weight": "model-00012-of-00024.safetensors",
284
+ "model.layers.16.self_attn.o_proj.weight": "model-00012-of-00024.safetensors",
285
+ "model.layers.16.self_attn.q_proj.weight": "model-00012-of-00024.safetensors",
286
+ "model.layers.16.self_attn.v_proj.weight": "model-00012-of-00024.safetensors",
287
+ "model.layers.17.block_sparse_moe.experts.0.w1.weight": "model-00013-of-00024.safetensors",
288
+ "model.layers.17.block_sparse_moe.experts.0.w2.weight": "model-00013-of-00024.safetensors",
289
+ "model.layers.17.block_sparse_moe.experts.0.w3.weight": "model-00013-of-00024.safetensors",
290
+ "model.layers.17.block_sparse_moe.experts.1.w1.weight": "model-00013-of-00024.safetensors",
291
+ "model.layers.17.block_sparse_moe.experts.1.w2.weight": "model-00013-of-00024.safetensors",
292
+ "model.layers.17.block_sparse_moe.experts.1.w3.weight": "model-00013-of-00024.safetensors",
293
+ "model.layers.17.block_sparse_moe.experts.2.w1.weight": "model-00013-of-00024.safetensors",
294
+ "model.layers.17.block_sparse_moe.experts.2.w2.weight": "model-00013-of-00024.safetensors",
295
+ "model.layers.17.block_sparse_moe.experts.2.w3.weight": "model-00013-of-00024.safetensors",
296
+ "model.layers.17.block_sparse_moe.experts.3.w1.weight": "model-00013-of-00024.safetensors",
297
+ "model.layers.17.block_sparse_moe.experts.3.w2.weight": "model-00013-of-00024.safetensors",
298
+ "model.layers.17.block_sparse_moe.experts.3.w3.weight": "model-00013-of-00024.safetensors",
299
+ "model.layers.17.block_sparse_moe.experts.4.w1.weight": "model-00013-of-00024.safetensors",
300
+ "model.layers.17.block_sparse_moe.experts.4.w2.weight": "model-00013-of-00024.safetensors",
301
+ "model.layers.17.block_sparse_moe.experts.4.w3.weight": "model-00014-of-00024.safetensors",
302
+ "model.layers.17.block_sparse_moe.experts.5.w1.weight": "model-00014-of-00024.safetensors",
303
+ "model.layers.17.block_sparse_moe.experts.5.w2.weight": "model-00014-of-00024.safetensors",
304
+ "model.layers.17.block_sparse_moe.experts.5.w3.weight": "model-00014-of-00024.safetensors",
305
+ "model.layers.17.block_sparse_moe.experts.6.w1.weight": "model-00014-of-00024.safetensors",
306
+ "model.layers.17.block_sparse_moe.experts.6.w2.weight": "model-00014-of-00024.safetensors",
307
+ "model.layers.17.block_sparse_moe.experts.6.w3.weight": "model-00014-of-00024.safetensors",
308
+ "model.layers.17.block_sparse_moe.experts.7.w1.weight": "model-00014-of-00024.safetensors",
309
+ "model.layers.17.block_sparse_moe.experts.7.w2.weight": "model-00014-of-00024.safetensors",
310
+ "model.layers.17.block_sparse_moe.experts.7.w3.weight": "model-00014-of-00024.safetensors",
311
+ "model.layers.17.block_sparse_moe.gate.weight": "model-00013-of-00024.safetensors",
312
+ "model.layers.17.input_layernorm.weight": "model-00014-of-00024.safetensors",
313
+ "model.layers.17.post_attention_layernorm.weight": "model-00014-of-00024.safetensors",
314
+ "model.layers.17.self_attn.k_proj.weight": "model-00013-of-00024.safetensors",
315
+ "model.layers.17.self_attn.o_proj.weight": "model-00013-of-00024.safetensors",
316
+ "model.layers.17.self_attn.q_proj.weight": "model-00013-of-00024.safetensors",
317
+ "model.layers.17.self_attn.v_proj.weight": "model-00013-of-00024.safetensors",
318
+ "model.layers.18.block_sparse_moe.experts.0.w1.weight": "model-00014-of-00024.safetensors",
319
+ "model.layers.18.block_sparse_moe.experts.0.w2.weight": "model-00014-of-00024.safetensors",
320
+ "model.layers.18.block_sparse_moe.experts.0.w3.weight": "model-00014-of-00024.safetensors",
321
+ "model.layers.18.block_sparse_moe.experts.1.w1.weight": "model-00014-of-00024.safetensors",
322
+ "model.layers.18.block_sparse_moe.experts.1.w2.weight": "model-00014-of-00024.safetensors",
323
+ "model.layers.18.block_sparse_moe.experts.1.w3.weight": "model-00014-of-00024.safetensors",
324
+ "model.layers.18.block_sparse_moe.experts.2.w1.weight": "model-00014-of-00024.safetensors",
325
+ "model.layers.18.block_sparse_moe.experts.2.w2.weight": "model-00014-of-00024.safetensors",
326
+ "model.layers.18.block_sparse_moe.experts.2.w3.weight": "model-00014-of-00024.safetensors",
327
+ "model.layers.18.block_sparse_moe.experts.3.w1.weight": "model-00014-of-00024.safetensors",
328
+ "model.layers.18.block_sparse_moe.experts.3.w2.weight": "model-00014-of-00024.safetensors",
329
+ "model.layers.18.block_sparse_moe.experts.3.w3.weight": "model-00014-of-00024.safetensors",
330
+ "model.layers.18.block_sparse_moe.experts.4.w1.weight": "model-00014-of-00024.safetensors",
331
+ "model.layers.18.block_sparse_moe.experts.4.w2.weight": "model-00014-of-00024.safetensors",
332
+ "model.layers.18.block_sparse_moe.experts.4.w3.weight": "model-00014-of-00024.safetensors",
333
+ "model.layers.18.block_sparse_moe.experts.5.w1.weight": "model-00014-of-00024.safetensors",
334
+ "model.layers.18.block_sparse_moe.experts.5.w2.weight": "model-00014-of-00024.safetensors",
335
+ "model.layers.18.block_sparse_moe.experts.5.w3.weight": "model-00014-of-00024.safetensors",
336
+ "model.layers.18.block_sparse_moe.experts.6.w1.weight": "model-00014-of-00024.safetensors",
337
+ "model.layers.18.block_sparse_moe.experts.6.w2.weight": "model-00014-of-00024.safetensors",
338
+ "model.layers.18.block_sparse_moe.experts.6.w3.weight": "model-00014-of-00024.safetensors",
339
+ "model.layers.18.block_sparse_moe.experts.7.w1.weight": "model-00014-of-00024.safetensors",
340
+ "model.layers.18.block_sparse_moe.experts.7.w2.weight": "model-00014-of-00024.safetensors",
341
+ "model.layers.18.block_sparse_moe.experts.7.w3.weight": "model-00015-of-00024.safetensors",
342
+ "model.layers.18.block_sparse_moe.gate.weight": "model-00014-of-00024.safetensors",
343
+ "model.layers.18.input_layernorm.weight": "model-00015-of-00024.safetensors",
344
+ "model.layers.18.post_attention_layernorm.weight": "model-00015-of-00024.safetensors",
345
+ "model.layers.18.self_attn.k_proj.weight": "model-00014-of-00024.safetensors",
346
+ "model.layers.18.self_attn.o_proj.weight": "model-00014-of-00024.safetensors",
347
+ "model.layers.18.self_attn.q_proj.weight": "model-00014-of-00024.safetensors",
348
+ "model.layers.18.self_attn.v_proj.weight": "model-00014-of-00024.safetensors",
349
+ "model.layers.19.block_sparse_moe.experts.0.w1.weight": "model-00015-of-00024.safetensors",
350
+ "model.layers.19.block_sparse_moe.experts.0.w2.weight": "model-00015-of-00024.safetensors",
351
+ "model.layers.19.block_sparse_moe.experts.0.w3.weight": "model-00015-of-00024.safetensors",
352
+ "model.layers.19.block_sparse_moe.experts.1.w1.weight": "model-00015-of-00024.safetensors",
353
+ "model.layers.19.block_sparse_moe.experts.1.w2.weight": "model-00015-of-00024.safetensors",
354
+ "model.layers.19.block_sparse_moe.experts.1.w3.weight": "model-00015-of-00024.safetensors",
355
+ "model.layers.19.block_sparse_moe.experts.2.w1.weight": "model-00015-of-00024.safetensors",
356
+ "model.layers.19.block_sparse_moe.experts.2.w2.weight": "model-00015-of-00024.safetensors",
357
+ "model.layers.19.block_sparse_moe.experts.2.w3.weight": "model-00015-of-00024.safetensors",
358
+ "model.layers.19.block_sparse_moe.experts.3.w1.weight": "model-00015-of-00024.safetensors",
359
+ "model.layers.19.block_sparse_moe.experts.3.w2.weight": "model-00015-of-00024.safetensors",
360
+ "model.layers.19.block_sparse_moe.experts.3.w3.weight": "model-00015-of-00024.safetensors",
361
+ "model.layers.19.block_sparse_moe.experts.4.w1.weight": "model-00015-of-00024.safetensors",
362
+ "model.layers.19.block_sparse_moe.experts.4.w2.weight": "model-00015-of-00024.safetensors",
363
+ "model.layers.19.block_sparse_moe.experts.4.w3.weight": "model-00015-of-00024.safetensors",
364
+ "model.layers.19.block_sparse_moe.experts.5.w1.weight": "model-00015-of-00024.safetensors",
365
+ "model.layers.19.block_sparse_moe.experts.5.w2.weight": "model-00015-of-00024.safetensors",
366
+ "model.layers.19.block_sparse_moe.experts.5.w3.weight": "model-00015-of-00024.safetensors",
367
+ "model.layers.19.block_sparse_moe.experts.6.w1.weight": "model-00015-of-00024.safetensors",
368
+ "model.layers.19.block_sparse_moe.experts.6.w2.weight": "model-00015-of-00024.safetensors",
369
+ "model.layers.19.block_sparse_moe.experts.6.w3.weight": "model-00015-of-00024.safetensors",
370
+ "model.layers.19.block_sparse_moe.experts.7.w1.weight": "model-00015-of-00024.safetensors",
371
+ "model.layers.19.block_sparse_moe.experts.7.w2.weight": "model-00015-of-00024.safetensors",
372
+ "model.layers.19.block_sparse_moe.experts.7.w3.weight": "model-00015-of-00024.safetensors",
373
+ "model.layers.19.block_sparse_moe.gate.weight": "model-00015-of-00024.safetensors",
374
+ "model.layers.19.input_layernorm.weight": "model-00015-of-00024.safetensors",
375
+ "model.layers.19.post_attention_layernorm.weight": "model-00015-of-00024.safetensors",
376
+ "model.layers.19.self_attn.k_proj.weight": "model-00015-of-00024.safetensors",
377
+ "model.layers.19.self_attn.o_proj.weight": "model-00015-of-00024.safetensors",
378
+ "model.layers.19.self_attn.q_proj.weight": "model-00015-of-00024.safetensors",
379
+ "model.layers.19.self_attn.v_proj.weight": "model-00015-of-00024.safetensors",
380
+ "model.layers.2.block_sparse_moe.experts.0.w1.weight": "model-00002-of-00024.safetensors",
381
+ "model.layers.2.block_sparse_moe.experts.0.w2.weight": "model-00002-of-00024.safetensors",
382
+ "model.layers.2.block_sparse_moe.experts.0.w3.weight": "model-00002-of-00024.safetensors",
383
+ "model.layers.2.block_sparse_moe.experts.1.w1.weight": "model-00002-of-00024.safetensors",
384
+ "model.layers.2.block_sparse_moe.experts.1.w2.weight": "model-00002-of-00024.safetensors",
385
+ "model.layers.2.block_sparse_moe.experts.1.w3.weight": "model-00002-of-00024.safetensors",
386
+ "model.layers.2.block_sparse_moe.experts.2.w1.weight": "model-00002-of-00024.safetensors",
387
+ "model.layers.2.block_sparse_moe.experts.2.w2.weight": "model-00002-of-00024.safetensors",
388
+ "model.layers.2.block_sparse_moe.experts.2.w3.weight": "model-00002-of-00024.safetensors",
389
+ "model.layers.2.block_sparse_moe.experts.3.w1.weight": "model-00002-of-00024.safetensors",
390
+ "model.layers.2.block_sparse_moe.experts.3.w2.weight": "model-00002-of-00024.safetensors",
391
+ "model.layers.2.block_sparse_moe.experts.3.w3.weight": "model-00002-of-00024.safetensors",
392
+ "model.layers.2.block_sparse_moe.experts.4.w1.weight": "model-00002-of-00024.safetensors",
393
+ "model.layers.2.block_sparse_moe.experts.4.w2.weight": "model-00002-of-00024.safetensors",
394
+ "model.layers.2.block_sparse_moe.experts.4.w3.weight": "model-00002-of-00024.safetensors",
395
+ "model.layers.2.block_sparse_moe.experts.5.w1.weight": "model-00003-of-00024.safetensors",
396
+ "model.layers.2.block_sparse_moe.experts.5.w2.weight": "model-00003-of-00024.safetensors",
397
+ "model.layers.2.block_sparse_moe.experts.5.w3.weight": "model-00003-of-00024.safetensors",
398
+ "model.layers.2.block_sparse_moe.experts.6.w1.weight": "model-00003-of-00024.safetensors",
399
+ "model.layers.2.block_sparse_moe.experts.6.w2.weight": "model-00003-of-00024.safetensors",
400
+ "model.layers.2.block_sparse_moe.experts.6.w3.weight": "model-00003-of-00024.safetensors",
401
+ "model.layers.2.block_sparse_moe.experts.7.w1.weight": "model-00003-of-00024.safetensors",
402
+ "model.layers.2.block_sparse_moe.experts.7.w2.weight": "model-00003-of-00024.safetensors",
403
+ "model.layers.2.block_sparse_moe.experts.7.w3.weight": "model-00003-of-00024.safetensors",
404
+ "model.layers.2.block_sparse_moe.gate.weight": "model-00002-of-00024.safetensors",
405
+ "model.layers.2.input_layernorm.weight": "model-00003-of-00024.safetensors",
406
+ "model.layers.2.post_attention_layernorm.weight": "model-00003-of-00024.safetensors",
407
+ "model.layers.2.self_attn.k_proj.weight": "model-00002-of-00024.safetensors",
408
+ "model.layers.2.self_attn.o_proj.weight": "model-00002-of-00024.safetensors",
409
+ "model.layers.2.self_attn.q_proj.weight": "model-00002-of-00024.safetensors",
410
+ "model.layers.2.self_attn.v_proj.weight": "model-00002-of-00024.safetensors",
411
+ "model.layers.20.block_sparse_moe.experts.0.w1.weight": "model-00015-of-00024.safetensors",
412
+ "model.layers.20.block_sparse_moe.experts.0.w2.weight": "model-00015-of-00024.safetensors",
413
+ "model.layers.20.block_sparse_moe.experts.0.w3.weight": "model-00015-of-00024.safetensors",
414
+ "model.layers.20.block_sparse_moe.experts.1.w1.weight": "model-00015-of-00024.safetensors",
415
+ "model.layers.20.block_sparse_moe.experts.1.w2.weight": "model-00015-of-00024.safetensors",
416
+ "model.layers.20.block_sparse_moe.experts.1.w3.weight": "model-00015-of-00024.safetensors",
417
+ "model.layers.20.block_sparse_moe.experts.2.w1.weight": "model-00015-of-00024.safetensors",
418
+ "model.layers.20.block_sparse_moe.experts.2.w2.weight": "model-00016-of-00024.safetensors",
419
+ "model.layers.20.block_sparse_moe.experts.2.w3.weight": "model-00016-of-00024.safetensors",
420
+ "model.layers.20.block_sparse_moe.experts.3.w1.weight": "model-00016-of-00024.safetensors",
421
+ "model.layers.20.block_sparse_moe.experts.3.w2.weight": "model-00016-of-00024.safetensors",
422
+ "model.layers.20.block_sparse_moe.experts.3.w3.weight": "model-00016-of-00024.safetensors",
423
+ "model.layers.20.block_sparse_moe.experts.4.w1.weight": "model-00016-of-00024.safetensors",
424
+ "model.layers.20.block_sparse_moe.experts.4.w2.weight": "model-00016-of-00024.safetensors",
425
+ "model.layers.20.block_sparse_moe.experts.4.w3.weight": "model-00016-of-00024.safetensors",
426
+ "model.layers.20.block_sparse_moe.experts.5.w1.weight": "model-00016-of-00024.safetensors",
427
+ "model.layers.20.block_sparse_moe.experts.5.w2.weight": "model-00016-of-00024.safetensors",
428
+ "model.layers.20.block_sparse_moe.experts.5.w3.weight": "model-00016-of-00024.safetensors",
429
+ "model.layers.20.block_sparse_moe.experts.6.w1.weight": "model-00016-of-00024.safetensors",
430
+ "model.layers.20.block_sparse_moe.experts.6.w2.weight": "model-00016-of-00024.safetensors",
431
+ "model.layers.20.block_sparse_moe.experts.6.w3.weight": "model-00016-of-00024.safetensors",
432
+ "model.layers.20.block_sparse_moe.experts.7.w1.weight": "model-00016-of-00024.safetensors",
433
+ "model.layers.20.block_sparse_moe.experts.7.w2.weight": "model-00016-of-00024.safetensors",
434
+ "model.layers.20.block_sparse_moe.experts.7.w3.weight": "model-00016-of-00024.safetensors",
435
+ "model.layers.20.block_sparse_moe.gate.weight": "model-00015-of-00024.safetensors",
436
+ "model.layers.20.input_layernorm.weight": "model-00016-of-00024.safetensors",
437
+ "model.layers.20.post_attention_layernorm.weight": "model-00016-of-00024.safetensors",
438
+ "model.layers.20.self_attn.k_proj.weight": "model-00015-of-00024.safetensors",
439
+ "model.layers.20.self_attn.o_proj.weight": "model-00015-of-00024.safetensors",
440
+ "model.layers.20.self_attn.q_proj.weight": "model-00015-of-00024.safetensors",
441
+ "model.layers.20.self_attn.v_proj.weight": "model-00015-of-00024.safetensors",
442
+ "model.layers.21.block_sparse_moe.experts.0.w1.weight": "model-00016-of-00024.safetensors",
443
+ "model.layers.21.block_sparse_moe.experts.0.w2.weight": "model-00016-of-00024.safetensors",
444
+ "model.layers.21.block_sparse_moe.experts.0.w3.weight": "model-00016-of-00024.safetensors",
445
+ "model.layers.21.block_sparse_moe.experts.1.w1.weight": "model-00016-of-00024.safetensors",
446
+ "model.layers.21.block_sparse_moe.experts.1.w2.weight": "model-00016-of-00024.safetensors",
447
+ "model.layers.21.block_sparse_moe.experts.1.w3.weight": "model-00016-of-00024.safetensors",
448
+ "model.layers.21.block_sparse_moe.experts.2.w1.weight": "model-00016-of-00024.safetensors",
449
+ "model.layers.21.block_sparse_moe.experts.2.w2.weight": "model-00016-of-00024.safetensors",
450
+ "model.layers.21.block_sparse_moe.experts.2.w3.weight": "model-00016-of-00024.safetensors",
451
+ "model.layers.21.block_sparse_moe.experts.3.w1.weight": "model-00016-of-00024.safetensors",
452
+ "model.layers.21.block_sparse_moe.experts.3.w2.weight": "model-00016-of-00024.safetensors",
453
+ "model.layers.21.block_sparse_moe.experts.3.w3.weight": "model-00016-of-00024.safetensors",
454
+ "model.layers.21.block_sparse_moe.experts.4.w1.weight": "model-00016-of-00024.safetensors",
455
+ "model.layers.21.block_sparse_moe.experts.4.w2.weight": "model-00016-of-00024.safetensors",
456
+ "model.layers.21.block_sparse_moe.experts.4.w3.weight": "model-00016-of-00024.safetensors",
457
+ "model.layers.21.block_sparse_moe.experts.5.w1.weight": "model-00016-of-00024.safetensors",
458
+ "model.layers.21.block_sparse_moe.experts.5.w2.weight": "model-00017-of-00024.safetensors",
459
+ "model.layers.21.block_sparse_moe.experts.5.w3.weight": "model-00017-of-00024.safetensors",
460
+ "model.layers.21.block_sparse_moe.experts.6.w1.weight": "model-00017-of-00024.safetensors",
461
+ "model.layers.21.block_sparse_moe.experts.6.w2.weight": "model-00017-of-00024.safetensors",
462
+ "model.layers.21.block_sparse_moe.experts.6.w3.weight": "model-00017-of-00024.safetensors",
463
+ "model.layers.21.block_sparse_moe.experts.7.w1.weight": "model-00017-of-00024.safetensors",
464
+ "model.layers.21.block_sparse_moe.experts.7.w2.weight": "model-00017-of-00024.safetensors",
465
+ "model.layers.21.block_sparse_moe.experts.7.w3.weight": "model-00017-of-00024.safetensors",
466
+ "model.layers.21.block_sparse_moe.gate.weight": "model-00016-of-00024.safetensors",
467
+ "model.layers.21.input_layernorm.weight": "model-00017-of-00024.safetensors",
468
+ "model.layers.21.post_attention_layernorm.weight": "model-00017-of-00024.safetensors",
469
+ "model.layers.21.self_attn.k_proj.weight": "model-00016-of-00024.safetensors",
470
+ "model.layers.21.self_attn.o_proj.weight": "model-00016-of-00024.safetensors",
471
+ "model.layers.21.self_attn.q_proj.weight": "model-00016-of-00024.safetensors",
472
+ "model.layers.21.self_attn.v_proj.weight": "model-00016-of-00024.safetensors",
473
+ "model.layers.22.block_sparse_moe.experts.0.w1.weight": "model-00017-of-00024.safetensors",
474
+ "model.layers.22.block_sparse_moe.experts.0.w2.weight": "model-00017-of-00024.safetensors",
475
+ "model.layers.22.block_sparse_moe.experts.0.w3.weight": "model-00017-of-00024.safetensors",
476
+ "model.layers.22.block_sparse_moe.experts.1.w1.weight": "model-00017-of-00024.safetensors",
477
+ "model.layers.22.block_sparse_moe.experts.1.w2.weight": "model-00017-of-00024.safetensors",
478
+ "model.layers.22.block_sparse_moe.experts.1.w3.weight": "model-00017-of-00024.safetensors",
479
+ "model.layers.22.block_sparse_moe.experts.2.w1.weight": "model-00017-of-00024.safetensors",
480
+ "model.layers.22.block_sparse_moe.experts.2.w2.weight": "model-00017-of-00024.safetensors",
481
+ "model.layers.22.block_sparse_moe.experts.2.w3.weight": "model-00017-of-00024.safetensors",
482
+ "model.layers.22.block_sparse_moe.experts.3.w1.weight": "model-00017-of-00024.safetensors",
483
+ "model.layers.22.block_sparse_moe.experts.3.w2.weight": "model-00017-of-00024.safetensors",
484
+ "model.layers.22.block_sparse_moe.experts.3.w3.weight": "model-00017-of-00024.safetensors",
485
+ "model.layers.22.block_sparse_moe.experts.4.w1.weight": "model-00017-of-00024.safetensors",
486
+ "model.layers.22.block_sparse_moe.experts.4.w2.weight": "model-00017-of-00024.safetensors",
487
+ "model.layers.22.block_sparse_moe.experts.4.w3.weight": "model-00017-of-00024.safetensors",
488
+ "model.layers.22.block_sparse_moe.experts.5.w1.weight": "model-00017-of-00024.safetensors",
489
+ "model.layers.22.block_sparse_moe.experts.5.w2.weight": "model-00017-of-00024.safetensors",
490
+ "model.layers.22.block_sparse_moe.experts.5.w3.weight": "model-00017-of-00024.safetensors",
491
+ "model.layers.22.block_sparse_moe.experts.6.w1.weight": "model-00017-of-00024.safetensors",
492
+ "model.layers.22.block_sparse_moe.experts.6.w2.weight": "model-00017-of-00024.safetensors",
493
+ "model.layers.22.block_sparse_moe.experts.6.w3.weight": "model-00017-of-00024.safetensors",
494
+ "model.layers.22.block_sparse_moe.experts.7.w1.weight": "model-00017-of-00024.safetensors",
495
+ "model.layers.22.block_sparse_moe.experts.7.w2.weight": "model-00017-of-00024.safetensors",
496
+ "model.layers.22.block_sparse_moe.experts.7.w3.weight": "model-00017-of-00024.safetensors",
497
+ "model.layers.22.block_sparse_moe.gate.weight": "model-00017-of-00024.safetensors",
498
+ "model.layers.22.input_layernorm.weight": "model-00017-of-00024.safetensors",
499
+ "model.layers.22.post_attention_layernorm.weight": "model-00017-of-00024.safetensors",
500
+ "model.layers.22.self_attn.k_proj.weight": "model-00017-of-00024.safetensors",
501
+ "model.layers.22.self_attn.o_proj.weight": "model-00017-of-00024.safetensors",
502
+ "model.layers.22.self_attn.q_proj.weight": "model-00017-of-00024.safetensors",
503
+ "model.layers.22.self_attn.v_proj.weight": "model-00017-of-00024.safetensors",
504
+ "model.layers.23.block_sparse_moe.experts.0.w1.weight": "model-00018-of-00024.safetensors",
505
+ "model.layers.23.block_sparse_moe.experts.0.w2.weight": "model-00018-of-00024.safetensors",
506
+ "model.layers.23.block_sparse_moe.experts.0.w3.weight": "model-00018-of-00024.safetensors",
507
+ "model.layers.23.block_sparse_moe.experts.1.w1.weight": "model-00018-of-00024.safetensors",
508
+ "model.layers.23.block_sparse_moe.experts.1.w2.weight": "model-00018-of-00024.safetensors",
509
+ "model.layers.23.block_sparse_moe.experts.1.w3.weight": "model-00018-of-00024.safetensors",
510
+ "model.layers.23.block_sparse_moe.experts.2.w1.weight": "model-00018-of-00024.safetensors",
511
+ "model.layers.23.block_sparse_moe.experts.2.w2.weight": "model-00018-of-00024.safetensors",
512
+ "model.layers.23.block_sparse_moe.experts.2.w3.weight": "model-00018-of-00024.safetensors",
513
+ "model.layers.23.block_sparse_moe.experts.3.w1.weight": "model-00018-of-00024.safetensors",
514
+ "model.layers.23.block_sparse_moe.experts.3.w2.weight": "model-00018-of-00024.safetensors",
515
+ "model.layers.23.block_sparse_moe.experts.3.w3.weight": "model-00018-of-00024.safetensors",
516
+ "model.layers.23.block_sparse_moe.experts.4.w1.weight": "model-00018-of-00024.safetensors",
517
+ "model.layers.23.block_sparse_moe.experts.4.w2.weight": "model-00018-of-00024.safetensors",
518
+ "model.layers.23.block_sparse_moe.experts.4.w3.weight": "model-00018-of-00024.safetensors",
519
+ "model.layers.23.block_sparse_moe.experts.5.w1.weight": "model-00018-of-00024.safetensors",
520
+ "model.layers.23.block_sparse_moe.experts.5.w2.weight": "model-00018-of-00024.safetensors",
521
+ "model.layers.23.block_sparse_moe.experts.5.w3.weight": "model-00018-of-00024.safetensors",
522
+ "model.layers.23.block_sparse_moe.experts.6.w1.weight": "model-00018-of-00024.safetensors",
523
+ "model.layers.23.block_sparse_moe.experts.6.w2.weight": "model-00018-of-00024.safetensors",
524
+ "model.layers.23.block_sparse_moe.experts.6.w3.weight": "model-00018-of-00024.safetensors",
525
+ "model.layers.23.block_sparse_moe.experts.7.w1.weight": "model-00018-of-00024.safetensors",
526
+ "model.layers.23.block_sparse_moe.experts.7.w2.weight": "model-00018-of-00024.safetensors",
527
+ "model.layers.23.block_sparse_moe.experts.7.w3.weight": "model-00018-of-00024.safetensors",
528
+ "model.layers.23.block_sparse_moe.gate.weight": "model-00017-of-00024.safetensors",
529
+ "model.layers.23.input_layernorm.weight": "model-00018-of-00024.safetensors",
530
+ "model.layers.23.post_attention_layernorm.weight": "model-00018-of-00024.safetensors",
531
+ "model.layers.23.self_attn.k_proj.weight": "model-00017-of-00024.safetensors",
532
+ "model.layers.23.self_attn.o_proj.weight": "model-00017-of-00024.safetensors",
533
+ "model.layers.23.self_attn.q_proj.weight": "model-00017-of-00024.safetensors",
534
+ "model.layers.23.self_attn.v_proj.weight": "model-00017-of-00024.safetensors",
535
+ "model.layers.24.block_sparse_moe.experts.0.w1.weight": "model-00018-of-00024.safetensors",
536
+ "model.layers.24.block_sparse_moe.experts.0.w2.weight": "model-00018-of-00024.safetensors",
537
+ "model.layers.24.block_sparse_moe.experts.0.w3.weight": "model-00018-of-00024.safetensors",
538
+ "model.layers.24.block_sparse_moe.experts.1.w1.weight": "model-00018-of-00024.safetensors",
539
+ "model.layers.24.block_sparse_moe.experts.1.w2.weight": "model-00018-of-00024.safetensors",
540
+ "model.layers.24.block_sparse_moe.experts.1.w3.weight": "model-00018-of-00024.safetensors",
541
+ "model.layers.24.block_sparse_moe.experts.2.w1.weight": "model-00018-of-00024.safetensors",
542
+ "model.layers.24.block_sparse_moe.experts.2.w2.weight": "model-00018-of-00024.safetensors",
543
+ "model.layers.24.block_sparse_moe.experts.2.w3.weight": "model-00018-of-00024.safetensors",
544
+ "model.layers.24.block_sparse_moe.experts.3.w1.weight": "model-00019-of-00024.safetensors",
545
+ "model.layers.24.block_sparse_moe.experts.3.w2.weight": "model-00019-of-00024.safetensors",
546
+ "model.layers.24.block_sparse_moe.experts.3.w3.weight": "model-00019-of-00024.safetensors",
547
+ "model.layers.24.block_sparse_moe.experts.4.w1.weight": "model-00019-of-00024.safetensors",
548
+ "model.layers.24.block_sparse_moe.experts.4.w2.weight": "model-00019-of-00024.safetensors",
549
+ "model.layers.24.block_sparse_moe.experts.4.w3.weight": "model-00019-of-00024.safetensors",
550
+ "model.layers.24.block_sparse_moe.experts.5.w1.weight": "model-00019-of-00024.safetensors",
551
+ "model.layers.24.block_sparse_moe.experts.5.w2.weight": "model-00019-of-00024.safetensors",
552
+ "model.layers.24.block_sparse_moe.experts.5.w3.weight": "model-00019-of-00024.safetensors",
553
+ "model.layers.24.block_sparse_moe.experts.6.w1.weight": "model-00019-of-00024.safetensors",
554
+ "model.layers.24.block_sparse_moe.experts.6.w2.weight": "model-00019-of-00024.safetensors",
555
+ "model.layers.24.block_sparse_moe.experts.6.w3.weight": "model-00019-of-00024.safetensors",
556
+ "model.layers.24.block_sparse_moe.experts.7.w1.weight": "model-00019-of-00024.safetensors",
557
+ "model.layers.24.block_sparse_moe.experts.7.w2.weight": "model-00019-of-00024.safetensors",
558
+ "model.layers.24.block_sparse_moe.experts.7.w3.weight": "model-00019-of-00024.safetensors",
559
+ "model.layers.24.block_sparse_moe.gate.weight": "model-00018-of-00024.safetensors",
560
+ "model.layers.24.input_layernorm.weight": "model-00019-of-00024.safetensors",
561
+ "model.layers.24.post_attention_layernorm.weight": "model-00019-of-00024.safetensors",
562
+ "model.layers.24.self_attn.k_proj.weight": "model-00018-of-00024.safetensors",
563
+ "model.layers.24.self_attn.o_proj.weight": "model-00018-of-00024.safetensors",
564
+ "model.layers.24.self_attn.q_proj.weight": "model-00018-of-00024.safetensors",
565
+ "model.layers.24.self_attn.v_proj.weight": "model-00018-of-00024.safetensors",
566
+ "model.layers.25.block_sparse_moe.experts.0.w1.weight": "model-00019-of-00024.safetensors",
567
+ "model.layers.25.block_sparse_moe.experts.0.w2.weight": "model-00019-of-00024.safetensors",
568
+ "model.layers.25.block_sparse_moe.experts.0.w3.weight": "model-00019-of-00024.safetensors",
569
+ "model.layers.25.block_sparse_moe.experts.1.w1.weight": "model-00019-of-00024.safetensors",
570
+ "model.layers.25.block_sparse_moe.experts.1.w2.weight": "model-00019-of-00024.safetensors",
571
+ "model.layers.25.block_sparse_moe.experts.1.w3.weight": "model-00019-of-00024.safetensors",
572
+ "model.layers.25.block_sparse_moe.experts.2.w1.weight": "model-00019-of-00024.safetensors",
573
+ "model.layers.25.block_sparse_moe.experts.2.w2.weight": "model-00019-of-00024.safetensors",
574
+ "model.layers.25.block_sparse_moe.experts.2.w3.weight": "model-00019-of-00024.safetensors",
575
+ "model.layers.25.block_sparse_moe.experts.3.w1.weight": "model-00019-of-00024.safetensors",
576
+ "model.layers.25.block_sparse_moe.experts.3.w2.weight": "model-00019-of-00024.safetensors",
577
+ "model.layers.25.block_sparse_moe.experts.3.w3.weight": "model-00019-of-00024.safetensors",
578
+ "model.layers.25.block_sparse_moe.experts.4.w1.weight": "model-00019-of-00024.safetensors",
579
+ "model.layers.25.block_sparse_moe.experts.4.w2.weight": "model-00019-of-00024.safetensors",
580
+ "model.layers.25.block_sparse_moe.experts.4.w3.weight": "model-00019-of-00024.safetensors",
581
+ "model.layers.25.block_sparse_moe.experts.5.w1.weight": "model-00019-of-00024.safetensors",
582
+ "model.layers.25.block_sparse_moe.experts.5.w2.weight": "model-00019-of-00024.safetensors",
583
+ "model.layers.25.block_sparse_moe.experts.5.w3.weight": "model-00019-of-00024.safetensors",
584
+ "model.layers.25.block_sparse_moe.experts.6.w1.weight": "model-00020-of-00024.safetensors",
585
+ "model.layers.25.block_sparse_moe.experts.6.w2.weight": "model-00020-of-00024.safetensors",
586
+ "model.layers.25.block_sparse_moe.experts.6.w3.weight": "model-00020-of-00024.safetensors",
587
+ "model.layers.25.block_sparse_moe.experts.7.w1.weight": "model-00020-of-00024.safetensors",
588
+ "model.layers.25.block_sparse_moe.experts.7.w2.weight": "model-00020-of-00024.safetensors",
589
+ "model.layers.25.block_sparse_moe.experts.7.w3.weight": "model-00020-of-00024.safetensors",
590
+ "model.layers.25.block_sparse_moe.gate.weight": "model-00019-of-00024.safetensors",
591
+ "model.layers.25.input_layernorm.weight": "model-00020-of-00024.safetensors",
592
+ "model.layers.25.post_attention_layernorm.weight": "model-00020-of-00024.safetensors",
593
+ "model.layers.25.self_attn.k_proj.weight": "model-00019-of-00024.safetensors",
594
+ "model.layers.25.self_attn.o_proj.weight": "model-00019-of-00024.safetensors",
595
+ "model.layers.25.self_attn.q_proj.weight": "model-00019-of-00024.safetensors",
596
+ "model.layers.25.self_attn.v_proj.weight": "model-00019-of-00024.safetensors",
597
+ "model.layers.26.block_sparse_moe.experts.0.w1.weight": "model-00020-of-00024.safetensors",
598
+ "model.layers.26.block_sparse_moe.experts.0.w2.weight": "model-00020-of-00024.safetensors",
599
+ "model.layers.26.block_sparse_moe.experts.0.w3.weight": "model-00020-of-00024.safetensors",
600
+ "model.layers.26.block_sparse_moe.experts.1.w1.weight": "model-00020-of-00024.safetensors",
601
+ "model.layers.26.block_sparse_moe.experts.1.w2.weight": "model-00020-of-00024.safetensors",
602
+ "model.layers.26.block_sparse_moe.experts.1.w3.weight": "model-00020-of-00024.safetensors",
603
+ "model.layers.26.block_sparse_moe.experts.2.w1.weight": "model-00020-of-00024.safetensors",
604
+ "model.layers.26.block_sparse_moe.experts.2.w2.weight": "model-00020-of-00024.safetensors",
605
+ "model.layers.26.block_sparse_moe.experts.2.w3.weight": "model-00020-of-00024.safetensors",
606
+ "model.layers.26.block_sparse_moe.experts.3.w1.weight": "model-00020-of-00024.safetensors",
607
+ "model.layers.26.block_sparse_moe.experts.3.w2.weight": "model-00020-of-00024.safetensors",
608
+ "model.layers.26.block_sparse_moe.experts.3.w3.weight": "model-00020-of-00024.safetensors",
609
+ "model.layers.26.block_sparse_moe.experts.4.w1.weight": "model-00020-of-00024.safetensors",
610
+ "model.layers.26.block_sparse_moe.experts.4.w2.weight": "model-00020-of-00024.safetensors",
611
+ "model.layers.26.block_sparse_moe.experts.4.w3.weight": "model-00020-of-00024.safetensors",
612
+ "model.layers.26.block_sparse_moe.experts.5.w1.weight": "model-00020-of-00024.safetensors",
613
+ "model.layers.26.block_sparse_moe.experts.5.w2.weight": "model-00020-of-00024.safetensors",
614
+ "model.layers.26.block_sparse_moe.experts.5.w3.weight": "model-00020-of-00024.safetensors",
615
+ "model.layers.26.block_sparse_moe.experts.6.w1.weight": "model-00020-of-00024.safetensors",
616
+ "model.layers.26.block_sparse_moe.experts.6.w2.weight": "model-00020-of-00024.safetensors",
617
+ "model.layers.26.block_sparse_moe.experts.6.w3.weight": "model-00020-of-00024.safetensors",
618
+ "model.layers.26.block_sparse_moe.experts.7.w1.weight": "model-00020-of-00024.safetensors",
619
+ "model.layers.26.block_sparse_moe.experts.7.w2.weight": "model-00020-of-00024.safetensors",
620
+ "model.layers.26.block_sparse_moe.experts.7.w3.weight": "model-00020-of-00024.safetensors",
621
+ "model.layers.26.block_sparse_moe.gate.weight": "model-00020-of-00024.safetensors",
622
+ "model.layers.26.input_layernorm.weight": "model-00020-of-00024.safetensors",
623
+ "model.layers.26.post_attention_layernorm.weight": "model-00020-of-00024.safetensors",
624
+ "model.layers.26.self_attn.k_proj.weight": "model-00020-of-00024.safetensors",
625
+ "model.layers.26.self_attn.o_proj.weight": "model-00020-of-00024.safetensors",
626
+ "model.layers.26.self_attn.q_proj.weight": "model-00020-of-00024.safetensors",
627
+ "model.layers.26.self_attn.v_proj.weight": "model-00020-of-00024.safetensors",
628
+ "model.layers.27.block_sparse_moe.experts.0.w1.weight": "model-00020-of-00024.safetensors",
629
+ "model.layers.27.block_sparse_moe.experts.0.w2.weight": "model-00020-of-00024.safetensors",
630
+ "model.layers.27.block_sparse_moe.experts.0.w3.weight": "model-00021-of-00024.safetensors",
631
+ "model.layers.27.block_sparse_moe.experts.1.w1.weight": "model-00021-of-00024.safetensors",
632
+ "model.layers.27.block_sparse_moe.experts.1.w2.weight": "model-00021-of-00024.safetensors",
633
+ "model.layers.27.block_sparse_moe.experts.1.w3.weight": "model-00021-of-00024.safetensors",
634
+ "model.layers.27.block_sparse_moe.experts.2.w1.weight": "model-00021-of-00024.safetensors",
635
+ "model.layers.27.block_sparse_moe.experts.2.w2.weight": "model-00021-of-00024.safetensors",
636
+ "model.layers.27.block_sparse_moe.experts.2.w3.weight": "model-00021-of-00024.safetensors",
637
+ "model.layers.27.block_sparse_moe.experts.3.w1.weight": "model-00021-of-00024.safetensors",
638
+ "model.layers.27.block_sparse_moe.experts.3.w2.weight": "model-00021-of-00024.safetensors",
639
+ "model.layers.27.block_sparse_moe.experts.3.w3.weight": "model-00021-of-00024.safetensors",
640
+ "model.layers.27.block_sparse_moe.experts.4.w1.weight": "model-00021-of-00024.safetensors",
641
+ "model.layers.27.block_sparse_moe.experts.4.w2.weight": "model-00021-of-00024.safetensors",
642
+ "model.layers.27.block_sparse_moe.experts.4.w3.weight": "model-00021-of-00024.safetensors",
643
+ "model.layers.27.block_sparse_moe.experts.5.w1.weight": "model-00021-of-00024.safetensors",
644
+ "model.layers.27.block_sparse_moe.experts.5.w2.weight": "model-00021-of-00024.safetensors",
645
+ "model.layers.27.block_sparse_moe.experts.5.w3.weight": "model-00021-of-00024.safetensors",
646
+ "model.layers.27.block_sparse_moe.experts.6.w1.weight": "model-00021-of-00024.safetensors",
647
+ "model.layers.27.block_sparse_moe.experts.6.w2.weight": "model-00021-of-00024.safetensors",
648
+ "model.layers.27.block_sparse_moe.experts.6.w3.weight": "model-00021-of-00024.safetensors",
649
+ "model.layers.27.block_sparse_moe.experts.7.w1.weight": "model-00021-of-00024.safetensors",
650
+ "model.layers.27.block_sparse_moe.experts.7.w2.weight": "model-00021-of-00024.safetensors",
651
+ "model.layers.27.block_sparse_moe.experts.7.w3.weight": "model-00021-of-00024.safetensors",
652
+ "model.layers.27.block_sparse_moe.gate.weight": "model-00020-of-00024.safetensors",
653
+ "model.layers.27.input_layernorm.weight": "model-00021-of-00024.safetensors",
654
+ "model.layers.27.post_attention_layernorm.weight": "model-00021-of-00024.safetensors",
655
+ "model.layers.27.self_attn.k_proj.weight": "model-00020-of-00024.safetensors",
656
+ "model.layers.27.self_attn.o_proj.weight": "model-00020-of-00024.safetensors",
657
+ "model.layers.27.self_attn.q_proj.weight": "model-00020-of-00024.safetensors",
658
+ "model.layers.27.self_attn.v_proj.weight": "model-00020-of-00024.safetensors",
659
+ "model.layers.28.block_sparse_moe.experts.0.w1.weight": "model-00021-of-00024.safetensors",
660
+ "model.layers.28.block_sparse_moe.experts.0.w2.weight": "model-00021-of-00024.safetensors",
661
+ "model.layers.28.block_sparse_moe.experts.0.w3.weight": "model-00021-of-00024.safetensors",
662
+ "model.layers.28.block_sparse_moe.experts.1.w1.weight": "model-00021-of-00024.safetensors",
663
+ "model.layers.28.block_sparse_moe.experts.1.w2.weight": "model-00021-of-00024.safetensors",
664
+ "model.layers.28.block_sparse_moe.experts.1.w3.weight": "model-00021-of-00024.safetensors",
665
+ "model.layers.28.block_sparse_moe.experts.2.w1.weight": "model-00021-of-00024.safetensors",
666
+ "model.layers.28.block_sparse_moe.experts.2.w2.weight": "model-00021-of-00024.safetensors",
667
+ "model.layers.28.block_sparse_moe.experts.2.w3.weight": "model-00021-of-00024.safetensors",
668
+ "model.layers.28.block_sparse_moe.experts.3.w1.weight": "model-00021-of-00024.safetensors",
669
+ "model.layers.28.block_sparse_moe.experts.3.w2.weight": "model-00021-of-00024.safetensors",
670
+ "model.layers.28.block_sparse_moe.experts.3.w3.weight": "model-00022-of-00024.safetensors",
671
+ "model.layers.28.block_sparse_moe.experts.4.w1.weight": "model-00022-of-00024.safetensors",
672
+ "model.layers.28.block_sparse_moe.experts.4.w2.weight": "model-00022-of-00024.safetensors",
673
+ "model.layers.28.block_sparse_moe.experts.4.w3.weight": "model-00022-of-00024.safetensors",
674
+ "model.layers.28.block_sparse_moe.experts.5.w1.weight": "model-00022-of-00024.safetensors",
675
+ "model.layers.28.block_sparse_moe.experts.5.w2.weight": "model-00022-of-00024.safetensors",
676
+ "model.layers.28.block_sparse_moe.experts.5.w3.weight": "model-00022-of-00024.safetensors",
677
+ "model.layers.28.block_sparse_moe.experts.6.w1.weight": "model-00022-of-00024.safetensors",
678
+ "model.layers.28.block_sparse_moe.experts.6.w2.weight": "model-00022-of-00024.safetensors",
679
+ "model.layers.28.block_sparse_moe.experts.6.w3.weight": "model-00022-of-00024.safetensors",
680
+ "model.layers.28.block_sparse_moe.experts.7.w1.weight": "model-00022-of-00024.safetensors",
681
+ "model.layers.28.block_sparse_moe.experts.7.w2.weight": "model-00022-of-00024.safetensors",
682
+ "model.layers.28.block_sparse_moe.experts.7.w3.weight": "model-00022-of-00024.safetensors",
683
+ "model.layers.28.block_sparse_moe.gate.weight": "model-00021-of-00024.safetensors",
684
+ "model.layers.28.input_layernorm.weight": "model-00022-of-00024.safetensors",
685
+ "model.layers.28.post_attention_layernorm.weight": "model-00022-of-00024.safetensors",
686
+ "model.layers.28.self_attn.k_proj.weight": "model-00021-of-00024.safetensors",
687
+ "model.layers.28.self_attn.o_proj.weight": "model-00021-of-00024.safetensors",
688
+ "model.layers.28.self_attn.q_proj.weight": "model-00021-of-00024.safetensors",
689
+ "model.layers.28.self_attn.v_proj.weight": "model-00021-of-00024.safetensors",
690
+ "model.layers.29.block_sparse_moe.experts.0.w1.weight": "model-00022-of-00024.safetensors",
691
+ "model.layers.29.block_sparse_moe.experts.0.w2.weight": "model-00022-of-00024.safetensors",
692
+ "model.layers.29.block_sparse_moe.experts.0.w3.weight": "model-00022-of-00024.safetensors",
693
+ "model.layers.29.block_sparse_moe.experts.1.w1.weight": "model-00022-of-00024.safetensors",
694
+ "model.layers.29.block_sparse_moe.experts.1.w2.weight": "model-00022-of-00024.safetensors",
695
+ "model.layers.29.block_sparse_moe.experts.1.w3.weight": "model-00022-of-00024.safetensors",
696
+ "model.layers.29.block_sparse_moe.experts.2.w1.weight": "model-00022-of-00024.safetensors",
697
+ "model.layers.29.block_sparse_moe.experts.2.w2.weight": "model-00022-of-00024.safetensors",
698
+ "model.layers.29.block_sparse_moe.experts.2.w3.weight": "model-00022-of-00024.safetensors",
699
+ "model.layers.29.block_sparse_moe.experts.3.w1.weight": "model-00022-of-00024.safetensors",
700
+ "model.layers.29.block_sparse_moe.experts.3.w2.weight": "model-00022-of-00024.safetensors",
701
+ "model.layers.29.block_sparse_moe.experts.3.w3.weight": "model-00022-of-00024.safetensors",
702
+ "model.layers.29.block_sparse_moe.experts.4.w1.weight": "model-00022-of-00024.safetensors",
703
+ "model.layers.29.block_sparse_moe.experts.4.w2.weight": "model-00022-of-00024.safetensors",
704
+ "model.layers.29.block_sparse_moe.experts.4.w3.weight": "model-00022-of-00024.safetensors",
705
+ "model.layers.29.block_sparse_moe.experts.5.w1.weight": "model-00022-of-00024.safetensors",
706
+ "model.layers.29.block_sparse_moe.experts.5.w2.weight": "model-00022-of-00024.safetensors",
707
+ "model.layers.29.block_sparse_moe.experts.5.w3.weight": "model-00022-of-00024.safetensors",
708
+ "model.layers.29.block_sparse_moe.experts.6.w1.weight": "model-00022-of-00024.safetensors",
709
+ "model.layers.29.block_sparse_moe.experts.6.w2.weight": "model-00022-of-00024.safetensors",
710
+ "model.layers.29.block_sparse_moe.experts.6.w3.weight": "model-00023-of-00024.safetensors",
711
+ "model.layers.29.block_sparse_moe.experts.7.w1.weight": "model-00023-of-00024.safetensors",
712
+ "model.layers.29.block_sparse_moe.experts.7.w2.weight": "model-00023-of-00024.safetensors",
713
+ "model.layers.29.block_sparse_moe.experts.7.w3.weight": "model-00023-of-00024.safetensors",
714
+ "model.layers.29.block_sparse_moe.gate.weight": "model-00022-of-00024.safetensors",
715
+ "model.layers.29.input_layernorm.weight": "model-00023-of-00024.safetensors",
716
+ "model.layers.29.post_attention_layernorm.weight": "model-00023-of-00024.safetensors",
717
+ "model.layers.29.self_attn.k_proj.weight": "model-00022-of-00024.safetensors",
718
+ "model.layers.29.self_attn.o_proj.weight": "model-00022-of-00024.safetensors",
719
+ "model.layers.29.self_attn.q_proj.weight": "model-00022-of-00024.safetensors",
720
+ "model.layers.29.self_attn.v_proj.weight": "model-00022-of-00024.safetensors",
721
+ "model.layers.3.block_sparse_moe.experts.0.w1.weight": "model-00003-of-00024.safetensors",
722
+ "model.layers.3.block_sparse_moe.experts.0.w2.weight": "model-00003-of-00024.safetensors",
723
+ "model.layers.3.block_sparse_moe.experts.0.w3.weight": "model-00003-of-00024.safetensors",
724
+ "model.layers.3.block_sparse_moe.experts.1.w1.weight": "model-00003-of-00024.safetensors",
725
+ "model.layers.3.block_sparse_moe.experts.1.w2.weight": "model-00003-of-00024.safetensors",
726
+ "model.layers.3.block_sparse_moe.experts.1.w3.weight": "model-00003-of-00024.safetensors",
727
+ "model.layers.3.block_sparse_moe.experts.2.w1.weight": "model-00003-of-00024.safetensors",
728
+ "model.layers.3.block_sparse_moe.experts.2.w2.weight": "model-00003-of-00024.safetensors",
729
+ "model.layers.3.block_sparse_moe.experts.2.w3.weight": "model-00003-of-00024.safetensors",
730
+ "model.layers.3.block_sparse_moe.experts.3.w1.weight": "model-00003-of-00024.safetensors",
731
+ "model.layers.3.block_sparse_moe.experts.3.w2.weight": "model-00003-of-00024.safetensors",
732
+ "model.layers.3.block_sparse_moe.experts.3.w3.weight": "model-00003-of-00024.safetensors",
733
+ "model.layers.3.block_sparse_moe.experts.4.w1.weight": "model-00003-of-00024.safetensors",
734
+ "model.layers.3.block_sparse_moe.experts.4.w2.weight": "model-00003-of-00024.safetensors",
735
+ "model.layers.3.block_sparse_moe.experts.4.w3.weight": "model-00003-of-00024.safetensors",
736
+ "model.layers.3.block_sparse_moe.experts.5.w1.weight": "model-00003-of-00024.safetensors",
737
+ "model.layers.3.block_sparse_moe.experts.5.w2.weight": "model-00003-of-00024.safetensors",
738
+ "model.layers.3.block_sparse_moe.experts.5.w3.weight": "model-00003-of-00024.safetensors",
739
+ "model.layers.3.block_sparse_moe.experts.6.w1.weight": "model-00003-of-00024.safetensors",
740
+ "model.layers.3.block_sparse_moe.experts.6.w2.weight": "model-00003-of-00024.safetensors",
741
+ "model.layers.3.block_sparse_moe.experts.6.w3.weight": "model-00003-of-00024.safetensors",
742
+ "model.layers.3.block_sparse_moe.experts.7.w1.weight": "model-00003-of-00024.safetensors",
743
+ "model.layers.3.block_sparse_moe.experts.7.w2.weight": "model-00003-of-00024.safetensors",
744
+ "model.layers.3.block_sparse_moe.experts.7.w3.weight": "model-00003-of-00024.safetensors",
745
+ "model.layers.3.block_sparse_moe.gate.weight": "model-00003-of-00024.safetensors",
746
+ "model.layers.3.input_layernorm.weight": "model-00003-of-00024.safetensors",
747
+ "model.layers.3.post_attention_layernorm.weight": "model-00003-of-00024.safetensors",
748
+ "model.layers.3.self_attn.k_proj.weight": "model-00003-of-00024.safetensors",
749
+ "model.layers.3.self_attn.o_proj.weight": "model-00003-of-00024.safetensors",
750
+ "model.layers.3.self_attn.q_proj.weight": "model-00003-of-00024.safetensors",
751
+ "model.layers.3.self_attn.v_proj.weight": "model-00003-of-00024.safetensors",
752
+ "model.layers.30.block_sparse_moe.experts.0.w1.weight": "model-00023-of-00024.safetensors",
753
+ "model.layers.30.block_sparse_moe.experts.0.w2.weight": "model-00023-of-00024.safetensors",
754
+ "model.layers.30.block_sparse_moe.experts.0.w3.weight": "model-00023-of-00024.safetensors",
755
+ "model.layers.30.block_sparse_moe.experts.1.w1.weight": "model-00023-of-00024.safetensors",
756
+ "model.layers.30.block_sparse_moe.experts.1.w2.weight": "model-00023-of-00024.safetensors",
757
+ "model.layers.30.block_sparse_moe.experts.1.w3.weight": "model-00023-of-00024.safetensors",
758
+ "model.layers.30.block_sparse_moe.experts.2.w1.weight": "model-00023-of-00024.safetensors",
759
+ "model.layers.30.block_sparse_moe.experts.2.w2.weight": "model-00023-of-00024.safetensors",
760
+ "model.layers.30.block_sparse_moe.experts.2.w3.weight": "model-00023-of-00024.safetensors",
761
+ "model.layers.30.block_sparse_moe.experts.3.w1.weight": "model-00023-of-00024.safetensors",
762
+ "model.layers.30.block_sparse_moe.experts.3.w2.weight": "model-00023-of-00024.safetensors",
763
+ "model.layers.30.block_sparse_moe.experts.3.w3.weight": "model-00023-of-00024.safetensors",
764
+ "model.layers.30.block_sparse_moe.experts.4.w1.weight": "model-00023-of-00024.safetensors",
765
+ "model.layers.30.block_sparse_moe.experts.4.w2.weight": "model-00023-of-00024.safetensors",
766
+ "model.layers.30.block_sparse_moe.experts.4.w3.weight": "model-00023-of-00024.safetensors",
767
+ "model.layers.30.block_sparse_moe.experts.5.w1.weight": "model-00023-of-00024.safetensors",
768
+ "model.layers.30.block_sparse_moe.experts.5.w2.weight": "model-00023-of-00024.safetensors",
769
+ "model.layers.30.block_sparse_moe.experts.5.w3.weight": "model-00023-of-00024.safetensors",
770
+ "model.layers.30.block_sparse_moe.experts.6.w1.weight": "model-00023-of-00024.safetensors",
771
+ "model.layers.30.block_sparse_moe.experts.6.w2.weight": "model-00023-of-00024.safetensors",
772
+ "model.layers.30.block_sparse_moe.experts.6.w3.weight": "model-00023-of-00024.safetensors",
773
+ "model.layers.30.block_sparse_moe.experts.7.w1.weight": "model-00023-of-00024.safetensors",
774
+ "model.layers.30.block_sparse_moe.experts.7.w2.weight": "model-00023-of-00024.safetensors",
775
+ "model.layers.30.block_sparse_moe.experts.7.w3.weight": "model-00023-of-00024.safetensors",
776
+ "model.layers.30.block_sparse_moe.gate.weight": "model-00023-of-00024.safetensors",
777
+ "model.layers.30.input_layernorm.weight": "model-00023-of-00024.safetensors",
778
+ "model.layers.30.post_attention_layernorm.weight": "model-00023-of-00024.safetensors",
779
+ "model.layers.30.self_attn.k_proj.weight": "model-00023-of-00024.safetensors",
780
+ "model.layers.30.self_attn.o_proj.weight": "model-00023-of-00024.safetensors",
781
+ "model.layers.30.self_attn.q_proj.weight": "model-00023-of-00024.safetensors",
782
+ "model.layers.30.self_attn.v_proj.weight": "model-00023-of-00024.safetensors",
783
+ "model.layers.31.block_sparse_moe.experts.0.w1.weight": "model-00023-of-00024.safetensors",
784
+ "model.layers.31.block_sparse_moe.experts.0.w2.weight": "model-00023-of-00024.safetensors",
785
+ "model.layers.31.block_sparse_moe.experts.0.w3.weight": "model-00023-of-00024.safetensors",
786
+ "model.layers.31.block_sparse_moe.experts.1.w1.weight": "model-00023-of-00024.safetensors",
787
+ "model.layers.31.block_sparse_moe.experts.1.w2.weight": "model-00024-of-00024.safetensors",
788
+ "model.layers.31.block_sparse_moe.experts.1.w3.weight": "model-00024-of-00024.safetensors",
789
+ "model.layers.31.block_sparse_moe.experts.2.w1.weight": "model-00024-of-00024.safetensors",
790
+ "model.layers.31.block_sparse_moe.experts.2.w2.weight": "model-00024-of-00024.safetensors",
791
+ "model.layers.31.block_sparse_moe.experts.2.w3.weight": "model-00024-of-00024.safetensors",
792
+ "model.layers.31.block_sparse_moe.experts.3.w1.weight": "model-00024-of-00024.safetensors",
793
+ "model.layers.31.block_sparse_moe.experts.3.w2.weight": "model-00024-of-00024.safetensors",
794
+ "model.layers.31.block_sparse_moe.experts.3.w3.weight": "model-00024-of-00024.safetensors",
795
+ "model.layers.31.block_sparse_moe.experts.4.w1.weight": "model-00024-of-00024.safetensors",
796
+ "model.layers.31.block_sparse_moe.experts.4.w2.weight": "model-00024-of-00024.safetensors",
797
+ "model.layers.31.block_sparse_moe.experts.4.w3.weight": "model-00024-of-00024.safetensors",
798
+ "model.layers.31.block_sparse_moe.experts.5.w1.weight": "model-00024-of-00024.safetensors",
799
+ "model.layers.31.block_sparse_moe.experts.5.w2.weight": "model-00024-of-00024.safetensors",
800
+ "model.layers.31.block_sparse_moe.experts.5.w3.weight": "model-00024-of-00024.safetensors",
801
+ "model.layers.31.block_sparse_moe.experts.6.w1.weight": "model-00024-of-00024.safetensors",
802
+ "model.layers.31.block_sparse_moe.experts.6.w2.weight": "model-00024-of-00024.safetensors",
803
+ "model.layers.31.block_sparse_moe.experts.6.w3.weight": "model-00024-of-00024.safetensors",
804
+ "model.layers.31.block_sparse_moe.experts.7.w1.weight": "model-00024-of-00024.safetensors",
805
+ "model.layers.31.block_sparse_moe.experts.7.w2.weight": "model-00024-of-00024.safetensors",
806
+ "model.layers.31.block_sparse_moe.experts.7.w3.weight": "model-00024-of-00024.safetensors",
807
+ "model.layers.31.block_sparse_moe.gate.weight": "model-00023-of-00024.safetensors",
808
+ "model.layers.31.input_layernorm.weight": "model-00024-of-00024.safetensors",
809
+ "model.layers.31.post_attention_layernorm.weight": "model-00024-of-00024.safetensors",
810
+ "model.layers.31.self_attn.k_proj.weight": "model-00023-of-00024.safetensors",
811
+ "model.layers.31.self_attn.o_proj.weight": "model-00023-of-00024.safetensors",
812
+ "model.layers.31.self_attn.q_proj.weight": "model-00023-of-00024.safetensors",
813
+ "model.layers.31.self_attn.v_proj.weight": "model-00023-of-00024.safetensors",
814
+ "model.layers.4.block_sparse_moe.experts.0.w1.weight": "model-00004-of-00024.safetensors",
815
+ "model.layers.4.block_sparse_moe.experts.0.w2.weight": "model-00004-of-00024.safetensors",
816
+ "model.layers.4.block_sparse_moe.experts.0.w3.weight": "model-00004-of-00024.safetensors",
817
+ "model.layers.4.block_sparse_moe.experts.1.w1.weight": "model-00004-of-00024.safetensors",
818
+ "model.layers.4.block_sparse_moe.experts.1.w2.weight": "model-00004-of-00024.safetensors",
819
+ "model.layers.4.block_sparse_moe.experts.1.w3.weight": "model-00004-of-00024.safetensors",
820
+ "model.layers.4.block_sparse_moe.experts.2.w1.weight": "model-00004-of-00024.safetensors",
821
+ "model.layers.4.block_sparse_moe.experts.2.w2.weight": "model-00004-of-00024.safetensors",
822
+ "model.layers.4.block_sparse_moe.experts.2.w3.weight": "model-00004-of-00024.safetensors",
823
+ "model.layers.4.block_sparse_moe.experts.3.w1.weight": "model-00004-of-00024.safetensors",
824
+ "model.layers.4.block_sparse_moe.experts.3.w2.weight": "model-00004-of-00024.safetensors",
825
+ "model.layers.4.block_sparse_moe.experts.3.w3.weight": "model-00004-of-00024.safetensors",
826
+ "model.layers.4.block_sparse_moe.experts.4.w1.weight": "model-00004-of-00024.safetensors",
827
+ "model.layers.4.block_sparse_moe.experts.4.w2.weight": "model-00004-of-00024.safetensors",
828
+ "model.layers.4.block_sparse_moe.experts.4.w3.weight": "model-00004-of-00024.safetensors",
829
+ "model.layers.4.block_sparse_moe.experts.5.w1.weight": "model-00004-of-00024.safetensors",
830
+ "model.layers.4.block_sparse_moe.experts.5.w2.weight": "model-00004-of-00024.safetensors",
831
+ "model.layers.4.block_sparse_moe.experts.5.w3.weight": "model-00004-of-00024.safetensors",
832
+ "model.layers.4.block_sparse_moe.experts.6.w1.weight": "model-00004-of-00024.safetensors",
833
+ "model.layers.4.block_sparse_moe.experts.6.w2.weight": "model-00004-of-00024.safetensors",
834
+ "model.layers.4.block_sparse_moe.experts.6.w3.weight": "model-00004-of-00024.safetensors",
835
+ "model.layers.4.block_sparse_moe.experts.7.w1.weight": "model-00004-of-00024.safetensors",
836
+ "model.layers.4.block_sparse_moe.experts.7.w2.weight": "model-00004-of-00024.safetensors",
837
+ "model.layers.4.block_sparse_moe.experts.7.w3.weight": "model-00004-of-00024.safetensors",
838
+ "model.layers.4.block_sparse_moe.gate.weight": "model-00004-of-00024.safetensors",
839
+ "model.layers.4.input_layernorm.weight": "model-00004-of-00024.safetensors",
840
+ "model.layers.4.post_attention_layernorm.weight": "model-00004-of-00024.safetensors",
841
+ "model.layers.4.self_attn.k_proj.weight": "model-00004-of-00024.safetensors",
842
+ "model.layers.4.self_attn.o_proj.weight": "model-00004-of-00024.safetensors",
843
+ "model.layers.4.self_attn.q_proj.weight": "model-00003-of-00024.safetensors",
844
+ "model.layers.4.self_attn.v_proj.weight": "model-00004-of-00024.safetensors",
845
+ "model.layers.5.block_sparse_moe.experts.0.w1.weight": "model-00004-of-00024.safetensors",
846
+ "model.layers.5.block_sparse_moe.experts.0.w2.weight": "model-00004-of-00024.safetensors",
847
+ "model.layers.5.block_sparse_moe.experts.0.w3.weight": "model-00004-of-00024.safetensors",
848
+ "model.layers.5.block_sparse_moe.experts.1.w1.weight": "model-00004-of-00024.safetensors",
849
+ "model.layers.5.block_sparse_moe.experts.1.w2.weight": "model-00004-of-00024.safetensors",
850
+ "model.layers.5.block_sparse_moe.experts.1.w3.weight": "model-00004-of-00024.safetensors",
851
+ "model.layers.5.block_sparse_moe.experts.2.w1.weight": "model-00004-of-00024.safetensors",
852
+ "model.layers.5.block_sparse_moe.experts.2.w2.weight": "model-00004-of-00024.safetensors",
853
+ "model.layers.5.block_sparse_moe.experts.2.w3.weight": "model-00005-of-00024.safetensors",
854
+ "model.layers.5.block_sparse_moe.experts.3.w1.weight": "model-00005-of-00024.safetensors",
855
+ "model.layers.5.block_sparse_moe.experts.3.w2.weight": "model-00005-of-00024.safetensors",
856
+ "model.layers.5.block_sparse_moe.experts.3.w3.weight": "model-00005-of-00024.safetensors",
857
+ "model.layers.5.block_sparse_moe.experts.4.w1.weight": "model-00005-of-00024.safetensors",
858
+ "model.layers.5.block_sparse_moe.experts.4.w2.weight": "model-00005-of-00024.safetensors",
859
+ "model.layers.5.block_sparse_moe.experts.4.w3.weight": "model-00005-of-00024.safetensors",
860
+ "model.layers.5.block_sparse_moe.experts.5.w1.weight": "model-00005-of-00024.safetensors",
861
+ "model.layers.5.block_sparse_moe.experts.5.w2.weight": "model-00005-of-00024.safetensors",
862
+ "model.layers.5.block_sparse_moe.experts.5.w3.weight": "model-00005-of-00024.safetensors",
863
+ "model.layers.5.block_sparse_moe.experts.6.w1.weight": "model-00005-of-00024.safetensors",
864
+ "model.layers.5.block_sparse_moe.experts.6.w2.weight": "model-00005-of-00024.safetensors",
865
+ "model.layers.5.block_sparse_moe.experts.6.w3.weight": "model-00005-of-00024.safetensors",
866
+ "model.layers.5.block_sparse_moe.experts.7.w1.weight": "model-00005-of-00024.safetensors",
867
+ "model.layers.5.block_sparse_moe.experts.7.w2.weight": "model-00005-of-00024.safetensors",
868
+ "model.layers.5.block_sparse_moe.experts.7.w3.weight": "model-00005-of-00024.safetensors",
869
+ "model.layers.5.block_sparse_moe.gate.weight": "model-00004-of-00024.safetensors",
870
+ "model.layers.5.input_layernorm.weight": "model-00005-of-00024.safetensors",
871
+ "model.layers.5.post_attention_layernorm.weight": "model-00005-of-00024.safetensors",
872
+ "model.layers.5.self_attn.k_proj.weight": "model-00004-of-00024.safetensors",
873
+ "model.layers.5.self_attn.o_proj.weight": "model-00004-of-00024.safetensors",
874
+ "model.layers.5.self_attn.q_proj.weight": "model-00004-of-00024.safetensors",
875
+ "model.layers.5.self_attn.v_proj.weight": "model-00004-of-00024.safetensors",
876
+ "model.layers.6.block_sparse_moe.experts.0.w1.weight": "model-00005-of-00024.safetensors",
877
+ "model.layers.6.block_sparse_moe.experts.0.w2.weight": "model-00005-of-00024.safetensors",
878
+ "model.layers.6.block_sparse_moe.experts.0.w3.weight": "model-00005-of-00024.safetensors",
879
+ "model.layers.6.block_sparse_moe.experts.1.w1.weight": "model-00005-of-00024.safetensors",
880
+ "model.layers.6.block_sparse_moe.experts.1.w2.weight": "model-00005-of-00024.safetensors",
881
+ "model.layers.6.block_sparse_moe.experts.1.w3.weight": "model-00005-of-00024.safetensors",
882
+ "model.layers.6.block_sparse_moe.experts.2.w1.weight": "model-00005-of-00024.safetensors",
883
+ "model.layers.6.block_sparse_moe.experts.2.w2.weight": "model-00005-of-00024.safetensors",
884
+ "model.layers.6.block_sparse_moe.experts.2.w3.weight": "model-00005-of-00024.safetensors",
885
+ "model.layers.6.block_sparse_moe.experts.3.w1.weight": "model-00005-of-00024.safetensors",
886
+ "model.layers.6.block_sparse_moe.experts.3.w2.weight": "model-00005-of-00024.safetensors",
887
+ "model.layers.6.block_sparse_moe.experts.3.w3.weight": "model-00005-of-00024.safetensors",
888
+ "model.layers.6.block_sparse_moe.experts.4.w1.weight": "model-00005-of-00024.safetensors",
889
+ "model.layers.6.block_sparse_moe.experts.4.w2.weight": "model-00005-of-00024.safetensors",
890
+ "model.layers.6.block_sparse_moe.experts.4.w3.weight": "model-00005-of-00024.safetensors",
891
+ "model.layers.6.block_sparse_moe.experts.5.w1.weight": "model-00005-of-00024.safetensors",
892
+ "model.layers.6.block_sparse_moe.experts.5.w2.weight": "model-00005-of-00024.safetensors",
893
+ "model.layers.6.block_sparse_moe.experts.5.w3.weight": "model-00006-of-00024.safetensors",
894
+ "model.layers.6.block_sparse_moe.experts.6.w1.weight": "model-00006-of-00024.safetensors",
895
+ "model.layers.6.block_sparse_moe.experts.6.w2.weight": "model-00006-of-00024.safetensors",
896
+ "model.layers.6.block_sparse_moe.experts.6.w3.weight": "model-00006-of-00024.safetensors",
897
+ "model.layers.6.block_sparse_moe.experts.7.w1.weight": "model-00006-of-00024.safetensors",
898
+ "model.layers.6.block_sparse_moe.experts.7.w2.weight": "model-00006-of-00024.safetensors",
899
+ "model.layers.6.block_sparse_moe.experts.7.w3.weight": "model-00006-of-00024.safetensors",
900
+ "model.layers.6.block_sparse_moe.gate.weight": "model-00005-of-00024.safetensors",
901
+ "model.layers.6.input_layernorm.weight": "model-00006-of-00024.safetensors",
902
+ "model.layers.6.post_attention_layernorm.weight": "model-00006-of-00024.safetensors",
903
+ "model.layers.6.self_attn.k_proj.weight": "model-00005-of-00024.safetensors",
904
+ "model.layers.6.self_attn.o_proj.weight": "model-00005-of-00024.safetensors",
905
+ "model.layers.6.self_attn.q_proj.weight": "model-00005-of-00024.safetensors",
906
+ "model.layers.6.self_attn.v_proj.weight": "model-00005-of-00024.safetensors",
907
+ "model.layers.7.block_sparse_moe.experts.0.w1.weight": "model-00006-of-00024.safetensors",
908
+ "model.layers.7.block_sparse_moe.experts.0.w2.weight": "model-00006-of-00024.safetensors",
909
+ "model.layers.7.block_sparse_moe.experts.0.w3.weight": "model-00006-of-00024.safetensors",
910
+ "model.layers.7.block_sparse_moe.experts.1.w1.weight": "model-00006-of-00024.safetensors",
911
+ "model.layers.7.block_sparse_moe.experts.1.w2.weight": "model-00006-of-00024.safetensors",
912
+ "model.layers.7.block_sparse_moe.experts.1.w3.weight": "model-00006-of-00024.safetensors",
913
+ "model.layers.7.block_sparse_moe.experts.2.w1.weight": "model-00006-of-00024.safetensors",
914
+ "model.layers.7.block_sparse_moe.experts.2.w2.weight": "model-00006-of-00024.safetensors",
915
+ "model.layers.7.block_sparse_moe.experts.2.w3.weight": "model-00006-of-00024.safetensors",
916
+ "model.layers.7.block_sparse_moe.experts.3.w1.weight": "model-00006-of-00024.safetensors",
917
+ "model.layers.7.block_sparse_moe.experts.3.w2.weight": "model-00006-of-00024.safetensors",
918
+ "model.layers.7.block_sparse_moe.experts.3.w3.weight": "model-00006-of-00024.safetensors",
919
+ "model.layers.7.block_sparse_moe.experts.4.w1.weight": "model-00006-of-00024.safetensors",
920
+ "model.layers.7.block_sparse_moe.experts.4.w2.weight": "model-00006-of-00024.safetensors",
921
+ "model.layers.7.block_sparse_moe.experts.4.w3.weight": "model-00006-of-00024.safetensors",
922
+ "model.layers.7.block_sparse_moe.experts.5.w1.weight": "model-00006-of-00024.safetensors",
923
+ "model.layers.7.block_sparse_moe.experts.5.w2.weight": "model-00006-of-00024.safetensors",
924
+ "model.layers.7.block_sparse_moe.experts.5.w3.weight": "model-00006-of-00024.safetensors",
925
+ "model.layers.7.block_sparse_moe.experts.6.w1.weight": "model-00006-of-00024.safetensors",
926
+ "model.layers.7.block_sparse_moe.experts.6.w2.weight": "model-00006-of-00024.safetensors",
927
+ "model.layers.7.block_sparse_moe.experts.6.w3.weight": "model-00006-of-00024.safetensors",
928
+ "model.layers.7.block_sparse_moe.experts.7.w1.weight": "model-00006-of-00024.safetensors",
929
+ "model.layers.7.block_sparse_moe.experts.7.w2.weight": "model-00006-of-00024.safetensors",
930
+ "model.layers.7.block_sparse_moe.experts.7.w3.weight": "model-00006-of-00024.safetensors",
931
+ "model.layers.7.block_sparse_moe.gate.weight": "model-00006-of-00024.safetensors",
932
+ "model.layers.7.input_layernorm.weight": "model-00006-of-00024.safetensors",
933
+ "model.layers.7.post_attention_layernorm.weight": "model-00006-of-00024.safetensors",
934
+ "model.layers.7.self_attn.k_proj.weight": "model-00006-of-00024.safetensors",
935
+ "model.layers.7.self_attn.o_proj.weight": "model-00006-of-00024.safetensors",
936
+ "model.layers.7.self_attn.q_proj.weight": "model-00006-of-00024.safetensors",
937
+ "model.layers.7.self_attn.v_proj.weight": "model-00006-of-00024.safetensors",
938
+ "model.layers.8.block_sparse_moe.experts.0.w1.weight": "model-00006-of-00024.safetensors",
939
+ "model.layers.8.block_sparse_moe.experts.0.w2.weight": "model-00007-of-00024.safetensors",
940
+ "model.layers.8.block_sparse_moe.experts.0.w3.weight": "model-00007-of-00024.safetensors",
941
+ "model.layers.8.block_sparse_moe.experts.1.w1.weight": "model-00007-of-00024.safetensors",
942
+ "model.layers.8.block_sparse_moe.experts.1.w2.weight": "model-00007-of-00024.safetensors",
943
+ "model.layers.8.block_sparse_moe.experts.1.w3.weight": "model-00007-of-00024.safetensors",
944
+ "model.layers.8.block_sparse_moe.experts.2.w1.weight": "model-00007-of-00024.safetensors",
945
+ "model.layers.8.block_sparse_moe.experts.2.w2.weight": "model-00007-of-00024.safetensors",
946
+ "model.layers.8.block_sparse_moe.experts.2.w3.weight": "model-00007-of-00024.safetensors",
947
+ "model.layers.8.block_sparse_moe.experts.3.w1.weight": "model-00007-of-00024.safetensors",
948
+ "model.layers.8.block_sparse_moe.experts.3.w2.weight": "model-00007-of-00024.safetensors",
949
+ "model.layers.8.block_sparse_moe.experts.3.w3.weight": "model-00007-of-00024.safetensors",
950
+ "model.layers.8.block_sparse_moe.experts.4.w1.weight": "model-00007-of-00024.safetensors",
951
+ "model.layers.8.block_sparse_moe.experts.4.w2.weight": "model-00007-of-00024.safetensors",
952
+ "model.layers.8.block_sparse_moe.experts.4.w3.weight": "model-00007-of-00024.safetensors",
953
+ "model.layers.8.block_sparse_moe.experts.5.w1.weight": "model-00007-of-00024.safetensors",
954
+ "model.layers.8.block_sparse_moe.experts.5.w2.weight": "model-00007-of-00024.safetensors",
955
+ "model.layers.8.block_sparse_moe.experts.5.w3.weight": "model-00007-of-00024.safetensors",
956
+ "model.layers.8.block_sparse_moe.experts.6.w1.weight": "model-00007-of-00024.safetensors",
957
+ "model.layers.8.block_sparse_moe.experts.6.w2.weight": "model-00007-of-00024.safetensors",
958
+ "model.layers.8.block_sparse_moe.experts.6.w3.weight": "model-00007-of-00024.safetensors",
959
+ "model.layers.8.block_sparse_moe.experts.7.w1.weight": "model-00007-of-00024.safetensors",
960
+ "model.layers.8.block_sparse_moe.experts.7.w2.weight": "model-00007-of-00024.safetensors",
961
+ "model.layers.8.block_sparse_moe.experts.7.w3.weight": "model-00007-of-00024.safetensors",
962
+ "model.layers.8.block_sparse_moe.gate.weight": "model-00006-of-00024.safetensors",
963
+ "model.layers.8.input_layernorm.weight": "model-00007-of-00024.safetensors",
964
+ "model.layers.8.post_attention_layernorm.weight": "model-00007-of-00024.safetensors",
965
+ "model.layers.8.self_attn.k_proj.weight": "model-00006-of-00024.safetensors",
966
+ "model.layers.8.self_attn.o_proj.weight": "model-00006-of-00024.safetensors",
967
+ "model.layers.8.self_attn.q_proj.weight": "model-00006-of-00024.safetensors",
968
+ "model.layers.8.self_attn.v_proj.weight": "model-00006-of-00024.safetensors",
969
+ "model.layers.9.block_sparse_moe.experts.0.w1.weight": "model-00007-of-00024.safetensors",
970
+ "model.layers.9.block_sparse_moe.experts.0.w2.weight": "model-00007-of-00024.safetensors",
971
+ "model.layers.9.block_sparse_moe.experts.0.w3.weight": "model-00007-of-00024.safetensors",
972
+ "model.layers.9.block_sparse_moe.experts.1.w1.weight": "model-00007-of-00024.safetensors",
973
+ "model.layers.9.block_sparse_moe.experts.1.w2.weight": "model-00007-of-00024.safetensors",
974
+ "model.layers.9.block_sparse_moe.experts.1.w3.weight": "model-00007-of-00024.safetensors",
975
+ "model.layers.9.block_sparse_moe.experts.2.w1.weight": "model-00007-of-00024.safetensors",
976
+ "model.layers.9.block_sparse_moe.experts.2.w2.weight": "model-00007-of-00024.safetensors",
977
+ "model.layers.9.block_sparse_moe.experts.2.w3.weight": "model-00007-of-00024.safetensors",
978
+ "model.layers.9.block_sparse_moe.experts.3.w1.weight": "model-00007-of-00024.safetensors",
979
+ "model.layers.9.block_sparse_moe.experts.3.w2.weight": "model-00008-of-00024.safetensors",
980
+ "model.layers.9.block_sparse_moe.experts.3.w3.weight": "model-00008-of-00024.safetensors",
981
+ "model.layers.9.block_sparse_moe.experts.4.w1.weight": "model-00008-of-00024.safetensors",
982
+ "model.layers.9.block_sparse_moe.experts.4.w2.weight": "model-00008-of-00024.safetensors",
983
+ "model.layers.9.block_sparse_moe.experts.4.w3.weight": "model-00008-of-00024.safetensors",
984
+ "model.layers.9.block_sparse_moe.experts.5.w1.weight": "model-00008-of-00024.safetensors",
985
+ "model.layers.9.block_sparse_moe.experts.5.w2.weight": "model-00008-of-00024.safetensors",
986
+ "model.layers.9.block_sparse_moe.experts.5.w3.weight": "model-00008-of-00024.safetensors",
987
+ "model.layers.9.block_sparse_moe.experts.6.w1.weight": "model-00008-of-00024.safetensors",
988
+ "model.layers.9.block_sparse_moe.experts.6.w2.weight": "model-00008-of-00024.safetensors",
989
+ "model.layers.9.block_sparse_moe.experts.6.w3.weight": "model-00008-of-00024.safetensors",
990
+ "model.layers.9.block_sparse_moe.experts.7.w1.weight": "model-00008-of-00024.safetensors",
991
+ "model.layers.9.block_sparse_moe.experts.7.w2.weight": "model-00008-of-00024.safetensors",
992
+ "model.layers.9.block_sparse_moe.experts.7.w3.weight": "model-00008-of-00024.safetensors",
993
+ "model.layers.9.block_sparse_moe.gate.weight": "model-00007-of-00024.safetensors",
994
+ "model.layers.9.input_layernorm.weight": "model-00008-of-00024.safetensors",
995
+ "model.layers.9.post_attention_layernorm.weight": "model-00008-of-00024.safetensors",
996
+ "model.layers.9.self_attn.k_proj.weight": "model-00007-of-00024.safetensors",
997
+ "model.layers.9.self_attn.o_proj.weight": "model-00007-of-00024.safetensors",
998
+ "model.layers.9.self_attn.q_proj.weight": "model-00007-of-00024.safetensors",
999
+ "model.layers.9.self_attn.v_proj.weight": "model-00007-of-00024.safetensors",
1000
+ "model.norm.weight": "model-00024-of-00024.safetensors"
1001
+ }
1002
+ }
output-00001-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26e3b2cf837bbd4f265fe562be18a5cf30d15396f631cb9101ef7f49774c2a7d
3
+ size 8578184920
output-00002-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5104f6011460a972542fa420456d1555586e8b0b76a8e0c97c5c473ad5811013
3
+ size 8581360056
output-00003-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e62dd751fb3d03cb160faa838ed57164d72dd445577d529656b6b669470e487d
3
+ size 6447659856
special_tokens_map.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "unk_token": {
17
+ "content": "<unk>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ }
23
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
3
+ size 493443
tokenizer_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ }
29
+ },
30
+ "additional_special_tokens": [],
31
+ "bos_token": "<s>",
32
+ "clean_up_tokenization_spaces": false,
33
+ "eos_token": "</s>",
34
+ "legacy": true,
35
+ "model_max_length": 1000000000000000019884624838656,
36
+ "pad_token": null,
37
+ "sp_model_kwargs": {},
38
+ "spaces_between_special_tokens": false,
39
+ "tokenizer_class": "LlamaTokenizer",
40
+ "unk_token": "<unk>",
41
+ "use_default_system_prompt": false,
42
+ "chat_template": "{%- for idx in range(0, messages|length) -%}\n{%- if messages[idx]['role'] == 'user' -%}\n{%- if idx > 1 -%}\n{{- bos_token + '[INST] ' + messages[idx]['content'] + ' [/INST]' -}}\n{%- else -%}\n{{- messages[idx]['content'] + ' [/INST]' -}}\n{%- endif -%}\n{% elif messages[idx]['role'] == 'system' %}\n{{- '[INST] <<SYS>>\\n' + messages[idx]['content'] + '\\n<</SYS>>\\n\\n' -}}\n{%- elif messages[idx]['role'] == 'assistant' -%}\n{{- ' ' + messages[idx]['content'] + ' ' + eos_token -}}\n{% endif %}\n{% endfor %}"
43
+ }