Upload folder using huggingface_hub
Browse files- README.md +120 -0
- config.json +36 -0
- convert_mistral_weights_to_hf-22B.py +290 -0
- generation_config.json +6 -0
- huggingface-metadata.txt +14 -0
- model.safetensors.index.json +514 -0
- output-00001-of-00002.safetensors +3 -0
- output-00002-of-00002.safetensors +3 -0
- special_tokens_map.json +23 -0
- tokenizer.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +42 -0
README.md
ADDED
@@ -0,0 +1,120 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
inference: false
|
3 |
+
license: other
|
4 |
+
license_name: mnpl
|
5 |
+
license_link: https://mistral.ai/licences/MNPL-0.1.md
|
6 |
+
tags:
|
7 |
+
- code
|
8 |
+
language:
|
9 |
+
- code
|
10 |
+
---
|
11 |
+
|
12 |
+
Converted using [this](https://huggingface.co/bullerwins/Codestral-22B-v0.1-hf/blob/main/convert_mistral_weights_to_hf-22B.py) script
|
13 |
+
|
14 |
+
# Model Card for Codestral-22B-v0.1
|
15 |
+
|
16 |
+
Codestrall-22B-v0.1 is trained on a diverse dataset of 80+ programming languages, including the most popular ones, such as Python, Java, C, C++, JavaScript, and Bash (more details in the [Blogpost](https://mistral.ai/news/codestral/)). The model can be queried:
|
17 |
+
- As instruct, for instance to answer any questions about a code snippet (write documentation, explain, factorize) or to generate code following specific indications
|
18 |
+
- As Fill in the Middle (FIM), to predict the middle tokens between a prefix and a suffix (very useful for software development add-ons like in VS Code)
|
19 |
+
|
20 |
+
|
21 |
+
## Installation
|
22 |
+
|
23 |
+
It is recommended to use `mistralai/Codestral-22B-v0.1` with [mistral-inference](https://github.com/mistralai/mistral-inference).
|
24 |
+
|
25 |
+
```
|
26 |
+
pip install mistral_inference
|
27 |
+
```
|
28 |
+
|
29 |
+
## Download
|
30 |
+
|
31 |
+
```py
|
32 |
+
from huggingface_hub import snapshot_download
|
33 |
+
from pathlib import Path
|
34 |
+
|
35 |
+
mistral_models_path = Path.home().joinpath('mistral_models', 'Codestral-22B-v0.1')
|
36 |
+
mistral_models_path.mkdir(parents=True, exist_ok=True)
|
37 |
+
|
38 |
+
snapshot_download(repo_id="mistralai/Codestral-22B-v0.1", allow_patterns=["params.json", "consolidated.safetensors", "tokenizer.model.v3"], local_dir=mistral_models_path)
|
39 |
+
```
|
40 |
+
|
41 |
+
### Chat
|
42 |
+
|
43 |
+
After installing `mistral_inference`, a `mistral-chat` CLI command should be available in your environment.
|
44 |
+
|
45 |
+
```
|
46 |
+
mistral-chat $HOME/mistral_models/Codestral-22B-v0.1 --instruct --max_tokens 256
|
47 |
+
```
|
48 |
+
|
49 |
+
Will generate an answer to "Write me a function that computes fibonacci in Rust" and should give something along the following lines:
|
50 |
+
|
51 |
+
```
|
52 |
+
Sure, here's a simple implementation of a function that computes the Fibonacci sequence in Rust. This function takes an integer `n` as an argument and returns the `n`th Fibonacci number.
|
53 |
+
|
54 |
+
fn fibonacci(n: u32) -> u32 {
|
55 |
+
match n {
|
56 |
+
0 => 0,
|
57 |
+
1 => 1,
|
58 |
+
_ => fibonacci(n - 1) + fibonacci(n - 2),
|
59 |
+
}
|
60 |
+
}
|
61 |
+
|
62 |
+
fn main() {
|
63 |
+
let n = 10;
|
64 |
+
println!("The {}th Fibonacci number is: {}", n, fibonacci(n));
|
65 |
+
}
|
66 |
+
|
67 |
+
This function uses recursion to calculate the Fibonacci number. However, it's not the most efficient solution because it performs a lot of redundant calculations. A more efficient solution would use a loop to iteratively calculate the Fibonacci numbers.
|
68 |
+
```
|
69 |
+
|
70 |
+
|
71 |
+
### Fill-in-the-middle (FIM)
|
72 |
+
|
73 |
+
After installing `mistral_inference` and running `pip install --upgrade mistral_common` to make sure to have mistral_common>=1.2 installed:
|
74 |
+
|
75 |
+
```py
|
76 |
+
from mistral_inference.model import Transformer
|
77 |
+
from mistral_inference.generate import generate
|
78 |
+
from mistral_common.tokens.tokenizers.mistral import MistralTokenizer
|
79 |
+
from mistral_common.tokens.instruct.request import FIMRequest
|
80 |
+
|
81 |
+
tokenizer = MistralTokenizer.v3()
|
82 |
+
model = Transformer.from_folder("~/codestral-22B-240529")
|
83 |
+
|
84 |
+
prefix = """def add("""
|
85 |
+
suffix = """ return sum"""
|
86 |
+
|
87 |
+
request = FIMRequest(prompt=prefix, suffix=suffix)
|
88 |
+
|
89 |
+
tokens = tokenizer.encode_fim(request).tokens
|
90 |
+
|
91 |
+
out_tokens, _ = generate([tokens], model, max_tokens=256, temperature=0.0, eos_id=tokenizer.instruct_tokenizer.tokenizer.eos_id)
|
92 |
+
result = tokenizer.decode(out_tokens[0])
|
93 |
+
|
94 |
+
middle = result.split(suffix)[0].strip()
|
95 |
+
print(middle)
|
96 |
+
```
|
97 |
+
|
98 |
+
Should give something along the following lines:
|
99 |
+
|
100 |
+
```
|
101 |
+
num1, num2):
|
102 |
+
|
103 |
+
# Add two numbers
|
104 |
+
sum = num1 + num2
|
105 |
+
|
106 |
+
# return the sum
|
107 |
+
```
|
108 |
+
|
109 |
+
## Limitations
|
110 |
+
|
111 |
+
The Codestral-22B-v0.1 does not have any moderation mechanisms. We're looking forward to engaging with the community on ways to
|
112 |
+
make the model finely respect guardrails, allowing for deployment in environments requiring moderated outputs.
|
113 |
+
|
114 |
+
## License
|
115 |
+
|
116 |
+
Codestral-22B-v0.1 is released under the `MNLP-0.1` license.
|
117 |
+
|
118 |
+
## The Mistral AI Team
|
119 |
+
|
120 |
+
Albert Jiang, Alexandre Sablayrolles, Alexis Tacnet, Antoine Roux, Arthur Mensch, Audrey Herblin-Stoop, Baptiste Bout, Baudouin de Monicault, Blanche Savary, Bam4d, Caroline Feldman, Devendra Singh Chaplot, Diego de las Casas, Eleonore Arcelin, Emma Bou Hanna, Etienne Metzger, Gianna Lengyel, Guillaume Bour, Guillaume Lample, Harizo Rajaona, Henri Roussez, Jean-Malo Delignon, Jia Li, Justus Murke, Kartik Khandelwal, Lawrence Stewart, Louis Martin, Louis Ternon, Lucile Saulnier, Lélio Renard Lavaud, Margaret Jennings, Marie Pellat, Marie Torelli, Marie-Anne Lachaux, Marjorie Janiewicz, Mickael Seznec, Nicolas Schuhl, Patrick von Platen, Romain Sauvestre, Pierre Stock, Sandeep Subramanian, Saurabh Garg, Sophia Yang, Szymon Antoniak, Teven Le Scao, Thibaut Lavril, Thibault Schueller, Timothée Lacroix, Théophile Gervet, Thomas Wang, Valera Nemychnikova, Wendy Shang, William El Sayed, William Marshall
|
config.json
ADDED
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"architectures": [
|
3 |
+
"MistralForCausalLM"
|
4 |
+
],
|
5 |
+
"attention_dropout": 0.0,
|
6 |
+
"bos_token_id": 1,
|
7 |
+
"eos_token_id": 2,
|
8 |
+
"hidden_act": "silu",
|
9 |
+
"hidden_size": 6144,
|
10 |
+
"initializer_range": 0.02,
|
11 |
+
"intermediate_size": 16384,
|
12 |
+
"max_position_embeddings": 32768,
|
13 |
+
"model_type": "mistral",
|
14 |
+
"num_attention_heads": 48,
|
15 |
+
"num_hidden_layers": 56,
|
16 |
+
"num_key_value_heads": 8,
|
17 |
+
"rms_norm_eps": 1e-05,
|
18 |
+
"rope_theta": 1000000.0,
|
19 |
+
"sliding_window": null,
|
20 |
+
"tie_word_embeddings": false,
|
21 |
+
"torch_dtype": "bfloat16",
|
22 |
+
"transformers_version": "4.40.2",
|
23 |
+
"use_cache": true,
|
24 |
+
"vocab_size": 32768,
|
25 |
+
"quantization_config": {
|
26 |
+
"quant_method": "exl2",
|
27 |
+
"version": "0.1.1",
|
28 |
+
"bits": 3.5,
|
29 |
+
"head_bits": 8,
|
30 |
+
"calibration": {
|
31 |
+
"rows": 100,
|
32 |
+
"length": 2048,
|
33 |
+
"dataset": "(default)"
|
34 |
+
}
|
35 |
+
}
|
36 |
+
}
|
convert_mistral_weights_to_hf-22B.py
ADDED
@@ -0,0 +1,290 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Copyright 2023 Mistral AI and The HuggingFace Inc. team. All rights reserved.
|
2 |
+
#
|
3 |
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
4 |
+
# you may not use this file except in compliance with the License.
|
5 |
+
# You may obtain a copy of the License at
|
6 |
+
#
|
7 |
+
# http://www.apache.org/licenses/LICENSE-2.0
|
8 |
+
#
|
9 |
+
# Unless required by applicable law or agreed to in writing, software
|
10 |
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
11 |
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
12 |
+
# See the License for the specific language governing permissions and
|
13 |
+
# limitations under the License.
|
14 |
+
import argparse
|
15 |
+
import gc
|
16 |
+
import json
|
17 |
+
import os
|
18 |
+
import shutil
|
19 |
+
import warnings
|
20 |
+
|
21 |
+
import torch
|
22 |
+
from safetensors.torch import load_file as safe_load_file
|
23 |
+
|
24 |
+
from transformers import (
|
25 |
+
LlamaTokenizer,
|
26 |
+
MistralConfig,
|
27 |
+
MistralForCausalLM,
|
28 |
+
)
|
29 |
+
|
30 |
+
|
31 |
+
try:
|
32 |
+
from transformers import LlamaTokenizerFast
|
33 |
+
|
34 |
+
tokenizer_class = LlamaTokenizerFast
|
35 |
+
except ImportError as e:
|
36 |
+
warnings.warn(e)
|
37 |
+
warnings.warn(
|
38 |
+
"The converted tokenizer will be the `slow` tokenizer. To use the fast, update your `tokenizers` library and re-run the tokenizer conversion"
|
39 |
+
)
|
40 |
+
tokenizer_class = LlamaTokenizer
|
41 |
+
|
42 |
+
"""
|
43 |
+
Sample usage:
|
44 |
+
|
45 |
+
```
|
46 |
+
python src/transformers/models/mistral/convert_mistral_weights_to_hf.py \
|
47 |
+
--input_dir /path/to/downloaded/mistral/weights --model_size 22B --output_dir /output/path
|
48 |
+
```
|
49 |
+
|
50 |
+
Thereafter, models can be loaded via:
|
51 |
+
|
52 |
+
```py
|
53 |
+
from transformers import MistralForCausalLM, LlamaTokenizer
|
54 |
+
|
55 |
+
model = MistralForCausalLM.from_pretrained("/output/path")
|
56 |
+
tokenizer = LlamaTokenizer.from_pretrained("/output/path")
|
57 |
+
```
|
58 |
+
|
59 |
+
Important note: you need to be able to host the whole model in RAM to execute this script (even if the biggest versions
|
60 |
+
come in several checkpoints they each contain a part of each weight of the model, so we need to load them all in RAM).
|
61 |
+
"""
|
62 |
+
|
63 |
+
NUM_SHARDS = {"22B": 1}
|
64 |
+
|
65 |
+
|
66 |
+
def compute_intermediate_size(n, ffn_dim_multiplier=1, multiple_of=256):
|
67 |
+
return multiple_of * ((int(ffn_dim_multiplier * int(8 * n / 3)) + multiple_of - 1) // multiple_of)
|
68 |
+
|
69 |
+
|
70 |
+
def read_json(path):
|
71 |
+
with open(path, "r") as f:
|
72 |
+
return json.load(f)
|
73 |
+
|
74 |
+
|
75 |
+
def write_json(text, path):
|
76 |
+
with open(path, "w") as f:
|
77 |
+
json.dump(text, f)
|
78 |
+
|
79 |
+
|
80 |
+
def write_model(model_path, input_base_path, model_size, tokenizer_path=None, safe_serialization=True, is_v3=False):
|
81 |
+
# for backward compatibility, before you needed the repo to be called `my_repo/model_size`
|
82 |
+
if not os.path.isfile(os.path.join(input_base_path, "params.json")):
|
83 |
+
input_base_path = os.path.join(input_base_path, model_size)
|
84 |
+
|
85 |
+
os.makedirs(model_path, exist_ok=True)
|
86 |
+
tmp_model_path = os.path.join(model_path, "tmp")
|
87 |
+
os.makedirs(tmp_model_path, exist_ok=True)
|
88 |
+
|
89 |
+
params = read_json(os.path.join(input_base_path, "params.json"))
|
90 |
+
num_shards = NUM_SHARDS[model_size]
|
91 |
+
|
92 |
+
sliding_window = params.get("sliding_window", None)
|
93 |
+
|
94 |
+
# For some reason this is a string in the params.json
|
95 |
+
if sliding_window is not None:
|
96 |
+
sliding_window = int(sliding_window)
|
97 |
+
|
98 |
+
n_layers = params["n_layers"]
|
99 |
+
n_heads = params["n_heads"]
|
100 |
+
n_heads_per_shard = n_heads // num_shards
|
101 |
+
dim = params["dim"]
|
102 |
+
dims_per_head = dim // n_heads
|
103 |
+
base = params.get("rope_theta", 10000.0)
|
104 |
+
inv_freq = 1.0 / (base ** (torch.arange(0, dims_per_head, 2).float() / dims_per_head))
|
105 |
+
max_position_embeddings = 4096 * 8
|
106 |
+
|
107 |
+
if tokenizer_path is not None:
|
108 |
+
tokenizer = tokenizer_class(tokenizer_path + ".v3" if is_v3 else "")
|
109 |
+
tokenizer.save_pretrained(model_path)
|
110 |
+
vocab_size = tokenizer.vocab_size if tokenizer_path is not None else 32000
|
111 |
+
|
112 |
+
if "n_kv_heads" in params:
|
113 |
+
num_key_value_heads = params["n_kv_heads"] # for GQA / MQA
|
114 |
+
num_local_key_value_heads = num_key_value_heads // num_shards
|
115 |
+
key_value_dim = dims_per_head * num_local_key_value_heads
|
116 |
+
else: # compatibility with other checkpoints
|
117 |
+
num_key_value_heads = n_heads
|
118 |
+
num_local_key_value_heads = n_heads_per_shard
|
119 |
+
key_value_dim = dim
|
120 |
+
|
121 |
+
# permute for sliced rotary
|
122 |
+
def permute(w, n_heads=n_heads, dim1=dim, dim2=dim):
|
123 |
+
return w.view(n_heads, dim1 // n_heads // 2, 2, dim2).transpose(1, 2).reshape(dim1, dim2)
|
124 |
+
|
125 |
+
print(f"Fetching all parameters from the checkpoint at {input_base_path}.")
|
126 |
+
|
127 |
+
# Load weights - for v3 models the consolidated weights are in a single file format in safetensors
|
128 |
+
if is_v3:
|
129 |
+
loaded = [safe_load_file(os.path.join(input_base_path, "consolidated.safetensors"))]
|
130 |
+
else:
|
131 |
+
loaded = [
|
132 |
+
torch.load(os.path.join(input_base_path, f"consolidated.{i:02d}.pth"), map_location="cpu")
|
133 |
+
for i in range(num_shards)
|
134 |
+
]
|
135 |
+
param_count = 0
|
136 |
+
index_dict = {"weight_map": {}}
|
137 |
+
for layer_i in range(n_layers):
|
138 |
+
filename = f"pytorch_model-{layer_i + 1}-of-{n_layers + 1}.bin"
|
139 |
+
|
140 |
+
# Sharded
|
141 |
+
# Note that attention.w{q,k,v,o}, feed_fordward.w[1,2,3], attention_norm.weight and ffn_norm.weight share
|
142 |
+
# the same storage object, saving attention_norm and ffn_norm will save other weights too, which is
|
143 |
+
# redundant as other weights will be stitched from multiple shards. To avoid that, they are cloned.
|
144 |
+
|
145 |
+
state_dict = {
|
146 |
+
f"model.layers.{layer_i}.input_layernorm.weight": loaded[0][
|
147 |
+
f"layers.{layer_i}.attention_norm.weight"
|
148 |
+
].clone(),
|
149 |
+
f"model.layers.{layer_i}.post_attention_layernorm.weight": loaded[0][
|
150 |
+
f"layers.{layer_i}.ffn_norm.weight"
|
151 |
+
].clone(),
|
152 |
+
}
|
153 |
+
state_dict[f"model.layers.{layer_i}.self_attn.q_proj.weight"] = permute(
|
154 |
+
torch.cat(
|
155 |
+
[
|
156 |
+
loaded[i][f"layers.{layer_i}.attention.wq.weight"].view(n_heads_per_shard, dims_per_head, dim)
|
157 |
+
for i in range(num_shards)
|
158 |
+
],
|
159 |
+
dim=0,
|
160 |
+
).reshape(dim, dim)
|
161 |
+
)
|
162 |
+
state_dict[f"model.layers.{layer_i}.self_attn.k_proj.weight"] = permute(
|
163 |
+
torch.cat(
|
164 |
+
[
|
165 |
+
loaded[i][f"layers.{layer_i}.attention.wk.weight"].view(
|
166 |
+
num_local_key_value_heads, dims_per_head, dim
|
167 |
+
)
|
168 |
+
for i in range(num_shards)
|
169 |
+
],
|
170 |
+
dim=0,
|
171 |
+
).reshape(key_value_dim, dim),
|
172 |
+
num_key_value_heads,
|
173 |
+
key_value_dim,
|
174 |
+
dim,
|
175 |
+
)
|
176 |
+
state_dict[f"model.layers.{layer_i}.self_attn.v_proj.weight"] = torch.cat(
|
177 |
+
[
|
178 |
+
loaded[i][f"layers.{layer_i}.attention.wv.weight"].view(num_local_key_value_heads, dims_per_head, dim)
|
179 |
+
for i in range(num_shards)
|
180 |
+
],
|
181 |
+
dim=0,
|
182 |
+
).reshape(key_value_dim, dim)
|
183 |
+
|
184 |
+
state_dict[f"model.layers.{layer_i}.self_attn.o_proj.weight"] = torch.cat(
|
185 |
+
[loaded[i][f"layers.{layer_i}.attention.wo.weight"] for i in range(num_shards)], dim=1
|
186 |
+
)
|
187 |
+
state_dict[f"model.layers.{layer_i}.mlp.gate_proj.weight"] = torch.cat(
|
188 |
+
[loaded[i][f"layers.{layer_i}.feed_forward.w1.weight"] for i in range(num_shards)], dim=0
|
189 |
+
)
|
190 |
+
state_dict[f"model.layers.{layer_i}.mlp.down_proj.weight"] = torch.cat(
|
191 |
+
[loaded[i][f"layers.{layer_i}.feed_forward.w2.weight"] for i in range(num_shards)], dim=1
|
192 |
+
)
|
193 |
+
state_dict[f"model.layers.{layer_i}.mlp.up_proj.weight"] = torch.cat(
|
194 |
+
[loaded[i][f"layers.{layer_i}.feed_forward.w3.weight"] for i in range(num_shards)], dim=0
|
195 |
+
)
|
196 |
+
|
197 |
+
state_dict[f"model.layers.{layer_i}.self_attn.rotary_emb.inv_freq"] = inv_freq
|
198 |
+
for k, v in state_dict.items():
|
199 |
+
index_dict["weight_map"][k] = filename
|
200 |
+
param_count += v.numel()
|
201 |
+
torch.save(state_dict, os.path.join(tmp_model_path, filename))
|
202 |
+
|
203 |
+
filename = f"pytorch_model-{n_layers + 1}-of-{n_layers + 1}.bin"
|
204 |
+
state_dict = {
|
205 |
+
"model.norm.weight": loaded[0]["norm.weight"],
|
206 |
+
"model.embed_tokens.weight": torch.cat([loaded[i]["tok_embeddings.weight"] for i in range(num_shards)], dim=1),
|
207 |
+
"lm_head.weight": torch.cat([loaded[i]["output.weight"] for i in range(num_shards)], dim=0),
|
208 |
+
}
|
209 |
+
|
210 |
+
for k, v in state_dict.items():
|
211 |
+
index_dict["weight_map"][k] = filename
|
212 |
+
param_count += v.numel()
|
213 |
+
torch.save(state_dict, os.path.join(tmp_model_path, filename))
|
214 |
+
|
215 |
+
# Write configs
|
216 |
+
index_dict["metadata"] = {"total_size": param_count * 2}
|
217 |
+
write_json(index_dict, os.path.join(tmp_model_path, "pytorch_model.bin.index.json"))
|
218 |
+
config = MistralConfig(
|
219 |
+
hidden_size=dim,
|
220 |
+
intermediate_size=params["hidden_dim"],
|
221 |
+
num_attention_heads=params["n_heads"],
|
222 |
+
num_hidden_layers=params["n_layers"],
|
223 |
+
rms_norm_eps=params["norm_eps"],
|
224 |
+
num_key_value_heads=num_key_value_heads,
|
225 |
+
vocab_size=vocab_size,
|
226 |
+
rope_theta=base,
|
227 |
+
max_position_embeddings=max_position_embeddings,
|
228 |
+
sliding_window=sliding_window,
|
229 |
+
)
|
230 |
+
config.save_pretrained(tmp_model_path)
|
231 |
+
|
232 |
+
# Make space so we can load the model properly now.
|
233 |
+
del state_dict
|
234 |
+
del loaded
|
235 |
+
gc.collect()
|
236 |
+
|
237 |
+
print("Loading the checkpoint in a Mistral model.")
|
238 |
+
model = MistralForCausalLM.from_pretrained(tmp_model_path, torch_dtype=torch.bfloat16, low_cpu_mem_usage=True)
|
239 |
+
# Avoid saving this as part of the config.
|
240 |
+
del model.config._name_or_path
|
241 |
+
model.config.torch_dtype = torch.float16
|
242 |
+
print("Saving in the Transformers format.")
|
243 |
+
|
244 |
+
model.save_pretrained(model_path, safe_serialization=safe_serialization)
|
245 |
+
shutil.rmtree(tmp_model_path)
|
246 |
+
|
247 |
+
|
248 |
+
def write_tokenizer(tokenizer_path, input_tokenizer_path):
|
249 |
+
# Initialize the tokenizer based on the `spm` model
|
250 |
+
print(f"Saving a {tokenizer_class.__name__} to {tokenizer_path}.")
|
251 |
+
tokenizer = tokenizer_class(input_tokenizer_path)
|
252 |
+
tokenizer.save_pretrained(tokenizer_path)
|
253 |
+
|
254 |
+
|
255 |
+
def main():
|
256 |
+
parser = argparse.ArgumentParser()
|
257 |
+
parser.add_argument(
|
258 |
+
"--input_dir",
|
259 |
+
help="Location of Mistral weights, which contains tokenizer.model and model folders",
|
260 |
+
)
|
261 |
+
parser.add_argument(
|
262 |
+
"--model_size",
|
263 |
+
choices=["22B", "tokenizer_only"],
|
264 |
+
help="'f' models correspond to the finetuned versions, and are specific to the Mistral2 official release. For more details on Mistral2, checkout the original repo: https://huggingface.co/meta-mistral",
|
265 |
+
)
|
266 |
+
parser.add_argument(
|
267 |
+
"--output_dir",
|
268 |
+
help="Location to write HF model and tokenizer",
|
269 |
+
)
|
270 |
+
parser.add_argument("--safe_serialization", type=bool, help="Whether or not to save using `safetensors`.")
|
271 |
+
parser.add_argument(
|
272 |
+
"--is_v3", action="store_true", help="Whether the checkpoints correspond to the 3rd version or not."
|
273 |
+
)
|
274 |
+
args = parser.parse_args()
|
275 |
+
spm_path = os.path.join(args.input_dir, "tokenizer.model")
|
276 |
+
if args.model_size != "tokenizer_only":
|
277 |
+
write_model(
|
278 |
+
model_path=args.output_dir,
|
279 |
+
input_base_path=args.input_dir,
|
280 |
+
model_size=args.model_size,
|
281 |
+
safe_serialization=args.safe_serialization,
|
282 |
+
tokenizer_path=spm_path,
|
283 |
+
is_v3=args.is_v3,
|
284 |
+
)
|
285 |
+
else:
|
286 |
+
write_tokenizer(args.output_dir, spm_path)
|
287 |
+
|
288 |
+
|
289 |
+
if __name__ == "__main__":
|
290 |
+
main()
|
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": "4.40.2"
|
6 |
+
}
|
huggingface-metadata.txt
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
url: https://huggingface.co/bullerwins/Codestral-22B-v0.1-hf
|
2 |
+
branch: main
|
3 |
+
download date: 2024-05-29 14:06:47
|
4 |
+
sha256sum:
|
5 |
+
6c911e66544527032c9e49f602ed0645f748045248eb8fb8ec9982866b899674 model-00001-of-00009.safetensors
|
6 |
+
cc1de07197a04eaeeaa6dcb7ed6604f729ed822e92273c25c112f85c366b5696 model-00002-of-00009.safetensors
|
7 |
+
90dc483e3b22d3d21a03edd588a8ffe5743b8dea33fc9f1ffc01eb1e529aedf8 model-00003-of-00009.safetensors
|
8 |
+
76ee31da7cdd8fde0a257030ffdf7d3fb293935a62b8469f6dec1c1a19e14eee model-00004-of-00009.safetensors
|
9 |
+
1d56824727ffaf568f7a1c7770fd5cb531df71ebe143567b1cb3968aca7f98cd model-00005-of-00009.safetensors
|
10 |
+
9f30bb3fdbcad8d1c00e0b421908bebc6cb5544669cd3c916ae592acb7263ae4 model-00006-of-00009.safetensors
|
11 |
+
54eb704485dce4f8c7c245169d25f394ea08dec1562a1ab981715f294ef93314 model-00007-of-00009.safetensors
|
12 |
+
cfbe26e02d475904ecc92cbe54a614607156aabed3503867d8af5023673d6374 model-00008-of-00009.safetensors
|
13 |
+
c834720ddae75dc683e52284ffe27ea35f48eb2c5500c71025925fe0dd398a8c model-00009-of-00009.safetensors
|
14 |
+
37f00374dea48658ee8f5d0f21895b9bc55cb0103939607c8185bfd1c6ca1f89 tokenizer.model
|
model.safetensors.index.json
ADDED
@@ -0,0 +1,514 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"metadata": {
|
3 |
+
"total_size": 44494565376
|
4 |
+
},
|
5 |
+
"weight_map": {
|
6 |
+
"lm_head.weight": "model-00009-of-00009.safetensors",
|
7 |
+
"model.embed_tokens.weight": "model-00001-of-00009.safetensors",
|
8 |
+
"model.layers.0.input_layernorm.weight": "model-00001-of-00009.safetensors",
|
9 |
+
"model.layers.0.mlp.down_proj.weight": "model-00001-of-00009.safetensors",
|
10 |
+
"model.layers.0.mlp.gate_proj.weight": "model-00001-of-00009.safetensors",
|
11 |
+
"model.layers.0.mlp.up_proj.weight": "model-00001-of-00009.safetensors",
|
12 |
+
"model.layers.0.post_attention_layernorm.weight": "model-00001-of-00009.safetensors",
|
13 |
+
"model.layers.0.self_attn.k_proj.weight": "model-00001-of-00009.safetensors",
|
14 |
+
"model.layers.0.self_attn.o_proj.weight": "model-00001-of-00009.safetensors",
|
15 |
+
"model.layers.0.self_attn.q_proj.weight": "model-00001-of-00009.safetensors",
|
16 |
+
"model.layers.0.self_attn.v_proj.weight": "model-00001-of-00009.safetensors",
|
17 |
+
"model.layers.1.input_layernorm.weight": "model-00001-of-00009.safetensors",
|
18 |
+
"model.layers.1.mlp.down_proj.weight": "model-00001-of-00009.safetensors",
|
19 |
+
"model.layers.1.mlp.gate_proj.weight": "model-00001-of-00009.safetensors",
|
20 |
+
"model.layers.1.mlp.up_proj.weight": "model-00001-of-00009.safetensors",
|
21 |
+
"model.layers.1.post_attention_layernorm.weight": "model-00001-of-00009.safetensors",
|
22 |
+
"model.layers.1.self_attn.k_proj.weight": "model-00001-of-00009.safetensors",
|
23 |
+
"model.layers.1.self_attn.o_proj.weight": "model-00001-of-00009.safetensors",
|
24 |
+
"model.layers.1.self_attn.q_proj.weight": "model-00001-of-00009.safetensors",
|
25 |
+
"model.layers.1.self_attn.v_proj.weight": "model-00001-of-00009.safetensors",
|
26 |
+
"model.layers.10.input_layernorm.weight": "model-00002-of-00009.safetensors",
|
27 |
+
"model.layers.10.mlp.down_proj.weight": "model-00002-of-00009.safetensors",
|
28 |
+
"model.layers.10.mlp.gate_proj.weight": "model-00002-of-00009.safetensors",
|
29 |
+
"model.layers.10.mlp.up_proj.weight": "model-00002-of-00009.safetensors",
|
30 |
+
"model.layers.10.post_attention_layernorm.weight": "model-00002-of-00009.safetensors",
|
31 |
+
"model.layers.10.self_attn.k_proj.weight": "model-00002-of-00009.safetensors",
|
32 |
+
"model.layers.10.self_attn.o_proj.weight": "model-00002-of-00009.safetensors",
|
33 |
+
"model.layers.10.self_attn.q_proj.weight": "model-00002-of-00009.safetensors",
|
34 |
+
"model.layers.10.self_attn.v_proj.weight": "model-00002-of-00009.safetensors",
|
35 |
+
"model.layers.11.input_layernorm.weight": "model-00002-of-00009.safetensors",
|
36 |
+
"model.layers.11.mlp.down_proj.weight": "model-00002-of-00009.safetensors",
|
37 |
+
"model.layers.11.mlp.gate_proj.weight": "model-00002-of-00009.safetensors",
|
38 |
+
"model.layers.11.mlp.up_proj.weight": "model-00002-of-00009.safetensors",
|
39 |
+
"model.layers.11.post_attention_layernorm.weight": "model-00002-of-00009.safetensors",
|
40 |
+
"model.layers.11.self_attn.k_proj.weight": "model-00002-of-00009.safetensors",
|
41 |
+
"model.layers.11.self_attn.o_proj.weight": "model-00002-of-00009.safetensors",
|
42 |
+
"model.layers.11.self_attn.q_proj.weight": "model-00002-of-00009.safetensors",
|
43 |
+
"model.layers.11.self_attn.v_proj.weight": "model-00002-of-00009.safetensors",
|
44 |
+
"model.layers.12.input_layernorm.weight": "model-00003-of-00009.safetensors",
|
45 |
+
"model.layers.12.mlp.down_proj.weight": "model-00003-of-00009.safetensors",
|
46 |
+
"model.layers.12.mlp.gate_proj.weight": "model-00003-of-00009.safetensors",
|
47 |
+
"model.layers.12.mlp.up_proj.weight": "model-00003-of-00009.safetensors",
|
48 |
+
"model.layers.12.post_attention_layernorm.weight": "model-00003-of-00009.safetensors",
|
49 |
+
"model.layers.12.self_attn.k_proj.weight": "model-00002-of-00009.safetensors",
|
50 |
+
"model.layers.12.self_attn.o_proj.weight": "model-00003-of-00009.safetensors",
|
51 |
+
"model.layers.12.self_attn.q_proj.weight": "model-00002-of-00009.safetensors",
|
52 |
+
"model.layers.12.self_attn.v_proj.weight": "model-00002-of-00009.safetensors",
|
53 |
+
"model.layers.13.input_layernorm.weight": "model-00003-of-00009.safetensors",
|
54 |
+
"model.layers.13.mlp.down_proj.weight": "model-00003-of-00009.safetensors",
|
55 |
+
"model.layers.13.mlp.gate_proj.weight": "model-00003-of-00009.safetensors",
|
56 |
+
"model.layers.13.mlp.up_proj.weight": "model-00003-of-00009.safetensors",
|
57 |
+
"model.layers.13.post_attention_layernorm.weight": "model-00003-of-00009.safetensors",
|
58 |
+
"model.layers.13.self_attn.k_proj.weight": "model-00003-of-00009.safetensors",
|
59 |
+
"model.layers.13.self_attn.o_proj.weight": "model-00003-of-00009.safetensors",
|
60 |
+
"model.layers.13.self_attn.q_proj.weight": "model-00003-of-00009.safetensors",
|
61 |
+
"model.layers.13.self_attn.v_proj.weight": "model-00003-of-00009.safetensors",
|
62 |
+
"model.layers.14.input_layernorm.weight": "model-00003-of-00009.safetensors",
|
63 |
+
"model.layers.14.mlp.down_proj.weight": "model-00003-of-00009.safetensors",
|
64 |
+
"model.layers.14.mlp.gate_proj.weight": "model-00003-of-00009.safetensors",
|
65 |
+
"model.layers.14.mlp.up_proj.weight": "model-00003-of-00009.safetensors",
|
66 |
+
"model.layers.14.post_attention_layernorm.weight": "model-00003-of-00009.safetensors",
|
67 |
+
"model.layers.14.self_attn.k_proj.weight": "model-00003-of-00009.safetensors",
|
68 |
+
"model.layers.14.self_attn.o_proj.weight": "model-00003-of-00009.safetensors",
|
69 |
+
"model.layers.14.self_attn.q_proj.weight": "model-00003-of-00009.safetensors",
|
70 |
+
"model.layers.14.self_attn.v_proj.weight": "model-00003-of-00009.safetensors",
|
71 |
+
"model.layers.15.input_layernorm.weight": "model-00003-of-00009.safetensors",
|
72 |
+
"model.layers.15.mlp.down_proj.weight": "model-00003-of-00009.safetensors",
|
73 |
+
"model.layers.15.mlp.gate_proj.weight": "model-00003-of-00009.safetensors",
|
74 |
+
"model.layers.15.mlp.up_proj.weight": "model-00003-of-00009.safetensors",
|
75 |
+
"model.layers.15.post_attention_layernorm.weight": "model-00003-of-00009.safetensors",
|
76 |
+
"model.layers.15.self_attn.k_proj.weight": "model-00003-of-00009.safetensors",
|
77 |
+
"model.layers.15.self_attn.o_proj.weight": "model-00003-of-00009.safetensors",
|
78 |
+
"model.layers.15.self_attn.q_proj.weight": "model-00003-of-00009.safetensors",
|
79 |
+
"model.layers.15.self_attn.v_proj.weight": "model-00003-of-00009.safetensors",
|
80 |
+
"model.layers.16.input_layernorm.weight": "model-00003-of-00009.safetensors",
|
81 |
+
"model.layers.16.mlp.down_proj.weight": "model-00003-of-00009.safetensors",
|
82 |
+
"model.layers.16.mlp.gate_proj.weight": "model-00003-of-00009.safetensors",
|
83 |
+
"model.layers.16.mlp.up_proj.weight": "model-00003-of-00009.safetensors",
|
84 |
+
"model.layers.16.post_attention_layernorm.weight": "model-00003-of-00009.safetensors",
|
85 |
+
"model.layers.16.self_attn.k_proj.weight": "model-00003-of-00009.safetensors",
|
86 |
+
"model.layers.16.self_attn.o_proj.weight": "model-00003-of-00009.safetensors",
|
87 |
+
"model.layers.16.self_attn.q_proj.weight": "model-00003-of-00009.safetensors",
|
88 |
+
"model.layers.16.self_attn.v_proj.weight": "model-00003-of-00009.safetensors",
|
89 |
+
"model.layers.17.input_layernorm.weight": "model-00003-of-00009.safetensors",
|
90 |
+
"model.layers.17.mlp.down_proj.weight": "model-00003-of-00009.safetensors",
|
91 |
+
"model.layers.17.mlp.gate_proj.weight": "model-00003-of-00009.safetensors",
|
92 |
+
"model.layers.17.mlp.up_proj.weight": "model-00003-of-00009.safetensors",
|
93 |
+
"model.layers.17.post_attention_layernorm.weight": "model-00003-of-00009.safetensors",
|
94 |
+
"model.layers.17.self_attn.k_proj.weight": "model-00003-of-00009.safetensors",
|
95 |
+
"model.layers.17.self_attn.o_proj.weight": "model-00003-of-00009.safetensors",
|
96 |
+
"model.layers.17.self_attn.q_proj.weight": "model-00003-of-00009.safetensors",
|
97 |
+
"model.layers.17.self_attn.v_proj.weight": "model-00003-of-00009.safetensors",
|
98 |
+
"model.layers.18.input_layernorm.weight": "model-00004-of-00009.safetensors",
|
99 |
+
"model.layers.18.mlp.down_proj.weight": "model-00004-of-00009.safetensors",
|
100 |
+
"model.layers.18.mlp.gate_proj.weight": "model-00003-of-00009.safetensors",
|
101 |
+
"model.layers.18.mlp.up_proj.weight": "model-00004-of-00009.safetensors",
|
102 |
+
"model.layers.18.post_attention_layernorm.weight": "model-00004-of-00009.safetensors",
|
103 |
+
"model.layers.18.self_attn.k_proj.weight": "model-00003-of-00009.safetensors",
|
104 |
+
"model.layers.18.self_attn.o_proj.weight": "model-00003-of-00009.safetensors",
|
105 |
+
"model.layers.18.self_attn.q_proj.weight": "model-00003-of-00009.safetensors",
|
106 |
+
"model.layers.18.self_attn.v_proj.weight": "model-00003-of-00009.safetensors",
|
107 |
+
"model.layers.19.input_layernorm.weight": "model-00004-of-00009.safetensors",
|
108 |
+
"model.layers.19.mlp.down_proj.weight": "model-00004-of-00009.safetensors",
|
109 |
+
"model.layers.19.mlp.gate_proj.weight": "model-00004-of-00009.safetensors",
|
110 |
+
"model.layers.19.mlp.up_proj.weight": "model-00004-of-00009.safetensors",
|
111 |
+
"model.layers.19.post_attention_layernorm.weight": "model-00004-of-00009.safetensors",
|
112 |
+
"model.layers.19.self_attn.k_proj.weight": "model-00004-of-00009.safetensors",
|
113 |
+
"model.layers.19.self_attn.o_proj.weight": "model-00004-of-00009.safetensors",
|
114 |
+
"model.layers.19.self_attn.q_proj.weight": "model-00004-of-00009.safetensors",
|
115 |
+
"model.layers.19.self_attn.v_proj.weight": "model-00004-of-00009.safetensors",
|
116 |
+
"model.layers.2.input_layernorm.weight": "model-00001-of-00009.safetensors",
|
117 |
+
"model.layers.2.mlp.down_proj.weight": "model-00001-of-00009.safetensors",
|
118 |
+
"model.layers.2.mlp.gate_proj.weight": "model-00001-of-00009.safetensors",
|
119 |
+
"model.layers.2.mlp.up_proj.weight": "model-00001-of-00009.safetensors",
|
120 |
+
"model.layers.2.post_attention_layernorm.weight": "model-00001-of-00009.safetensors",
|
121 |
+
"model.layers.2.self_attn.k_proj.weight": "model-00001-of-00009.safetensors",
|
122 |
+
"model.layers.2.self_attn.o_proj.weight": "model-00001-of-00009.safetensors",
|
123 |
+
"model.layers.2.self_attn.q_proj.weight": "model-00001-of-00009.safetensors",
|
124 |
+
"model.layers.2.self_attn.v_proj.weight": "model-00001-of-00009.safetensors",
|
125 |
+
"model.layers.20.input_layernorm.weight": "model-00004-of-00009.safetensors",
|
126 |
+
"model.layers.20.mlp.down_proj.weight": "model-00004-of-00009.safetensors",
|
127 |
+
"model.layers.20.mlp.gate_proj.weight": "model-00004-of-00009.safetensors",
|
128 |
+
"model.layers.20.mlp.up_proj.weight": "model-00004-of-00009.safetensors",
|
129 |
+
"model.layers.20.post_attention_layernorm.weight": "model-00004-of-00009.safetensors",
|
130 |
+
"model.layers.20.self_attn.k_proj.weight": "model-00004-of-00009.safetensors",
|
131 |
+
"model.layers.20.self_attn.o_proj.weight": "model-00004-of-00009.safetensors",
|
132 |
+
"model.layers.20.self_attn.q_proj.weight": "model-00004-of-00009.safetensors",
|
133 |
+
"model.layers.20.self_attn.v_proj.weight": "model-00004-of-00009.safetensors",
|
134 |
+
"model.layers.21.input_layernorm.weight": "model-00004-of-00009.safetensors",
|
135 |
+
"model.layers.21.mlp.down_proj.weight": "model-00004-of-00009.safetensors",
|
136 |
+
"model.layers.21.mlp.gate_proj.weight": "model-00004-of-00009.safetensors",
|
137 |
+
"model.layers.21.mlp.up_proj.weight": "model-00004-of-00009.safetensors",
|
138 |
+
"model.layers.21.post_attention_layernorm.weight": "model-00004-of-00009.safetensors",
|
139 |
+
"model.layers.21.self_attn.k_proj.weight": "model-00004-of-00009.safetensors",
|
140 |
+
"model.layers.21.self_attn.o_proj.weight": "model-00004-of-00009.safetensors",
|
141 |
+
"model.layers.21.self_attn.q_proj.weight": "model-00004-of-00009.safetensors",
|
142 |
+
"model.layers.21.self_attn.v_proj.weight": "model-00004-of-00009.safetensors",
|
143 |
+
"model.layers.22.input_layernorm.weight": "model-00004-of-00009.safetensors",
|
144 |
+
"model.layers.22.mlp.down_proj.weight": "model-00004-of-00009.safetensors",
|
145 |
+
"model.layers.22.mlp.gate_proj.weight": "model-00004-of-00009.safetensors",
|
146 |
+
"model.layers.22.mlp.up_proj.weight": "model-00004-of-00009.safetensors",
|
147 |
+
"model.layers.22.post_attention_layernorm.weight": "model-00004-of-00009.safetensors",
|
148 |
+
"model.layers.22.self_attn.k_proj.weight": "model-00004-of-00009.safetensors",
|
149 |
+
"model.layers.22.self_attn.o_proj.weight": "model-00004-of-00009.safetensors",
|
150 |
+
"model.layers.22.self_attn.q_proj.weight": "model-00004-of-00009.safetensors",
|
151 |
+
"model.layers.22.self_attn.v_proj.weight": "model-00004-of-00009.safetensors",
|
152 |
+
"model.layers.23.input_layernorm.weight": "model-00004-of-00009.safetensors",
|
153 |
+
"model.layers.23.mlp.down_proj.weight": "model-00004-of-00009.safetensors",
|
154 |
+
"model.layers.23.mlp.gate_proj.weight": "model-00004-of-00009.safetensors",
|
155 |
+
"model.layers.23.mlp.up_proj.weight": "model-00004-of-00009.safetensors",
|
156 |
+
"model.layers.23.post_attention_layernorm.weight": "model-00004-of-00009.safetensors",
|
157 |
+
"model.layers.23.self_attn.k_proj.weight": "model-00004-of-00009.safetensors",
|
158 |
+
"model.layers.23.self_attn.o_proj.weight": "model-00004-of-00009.safetensors",
|
159 |
+
"model.layers.23.self_attn.q_proj.weight": "model-00004-of-00009.safetensors",
|
160 |
+
"model.layers.23.self_attn.v_proj.weight": "model-00004-of-00009.safetensors",
|
161 |
+
"model.layers.24.input_layernorm.weight": "model-00005-of-00009.safetensors",
|
162 |
+
"model.layers.24.mlp.down_proj.weight": "model-00005-of-00009.safetensors",
|
163 |
+
"model.layers.24.mlp.gate_proj.weight": "model-00004-of-00009.safetensors",
|
164 |
+
"model.layers.24.mlp.up_proj.weight": "model-00004-of-00009.safetensors",
|
165 |
+
"model.layers.24.post_attention_layernorm.weight": "model-00005-of-00009.safetensors",
|
166 |
+
"model.layers.24.self_attn.k_proj.weight": "model-00004-of-00009.safetensors",
|
167 |
+
"model.layers.24.self_attn.o_proj.weight": "model-00004-of-00009.safetensors",
|
168 |
+
"model.layers.24.self_attn.q_proj.weight": "model-00004-of-00009.safetensors",
|
169 |
+
"model.layers.24.self_attn.v_proj.weight": "model-00004-of-00009.safetensors",
|
170 |
+
"model.layers.25.input_layernorm.weight": "model-00005-of-00009.safetensors",
|
171 |
+
"model.layers.25.mlp.down_proj.weight": "model-00005-of-00009.safetensors",
|
172 |
+
"model.layers.25.mlp.gate_proj.weight": "model-00005-of-00009.safetensors",
|
173 |
+
"model.layers.25.mlp.up_proj.weight": "model-00005-of-00009.safetensors",
|
174 |
+
"model.layers.25.post_attention_layernorm.weight": "model-00005-of-00009.safetensors",
|
175 |
+
"model.layers.25.self_attn.k_proj.weight": "model-00005-of-00009.safetensors",
|
176 |
+
"model.layers.25.self_attn.o_proj.weight": "model-00005-of-00009.safetensors",
|
177 |
+
"model.layers.25.self_attn.q_proj.weight": "model-00005-of-00009.safetensors",
|
178 |
+
"model.layers.25.self_attn.v_proj.weight": "model-00005-of-00009.safetensors",
|
179 |
+
"model.layers.26.input_layernorm.weight": "model-00005-of-00009.safetensors",
|
180 |
+
"model.layers.26.mlp.down_proj.weight": "model-00005-of-00009.safetensors",
|
181 |
+
"model.layers.26.mlp.gate_proj.weight": "model-00005-of-00009.safetensors",
|
182 |
+
"model.layers.26.mlp.up_proj.weight": "model-00005-of-00009.safetensors",
|
183 |
+
"model.layers.26.post_attention_layernorm.weight": "model-00005-of-00009.safetensors",
|
184 |
+
"model.layers.26.self_attn.k_proj.weight": "model-00005-of-00009.safetensors",
|
185 |
+
"model.layers.26.self_attn.o_proj.weight": "model-00005-of-00009.safetensors",
|
186 |
+
"model.layers.26.self_attn.q_proj.weight": "model-00005-of-00009.safetensors",
|
187 |
+
"model.layers.26.self_attn.v_proj.weight": "model-00005-of-00009.safetensors",
|
188 |
+
"model.layers.27.input_layernorm.weight": "model-00005-of-00009.safetensors",
|
189 |
+
"model.layers.27.mlp.down_proj.weight": "model-00005-of-00009.safetensors",
|
190 |
+
"model.layers.27.mlp.gate_proj.weight": "model-00005-of-00009.safetensors",
|
191 |
+
"model.layers.27.mlp.up_proj.weight": "model-00005-of-00009.safetensors",
|
192 |
+
"model.layers.27.post_attention_layernorm.weight": "model-00005-of-00009.safetensors",
|
193 |
+
"model.layers.27.self_attn.k_proj.weight": "model-00005-of-00009.safetensors",
|
194 |
+
"model.layers.27.self_attn.o_proj.weight": "model-00005-of-00009.safetensors",
|
195 |
+
"model.layers.27.self_attn.q_proj.weight": "model-00005-of-00009.safetensors",
|
196 |
+
"model.layers.27.self_attn.v_proj.weight": "model-00005-of-00009.safetensors",
|
197 |
+
"model.layers.28.input_layernorm.weight": "model-00005-of-00009.safetensors",
|
198 |
+
"model.layers.28.mlp.down_proj.weight": "model-00005-of-00009.safetensors",
|
199 |
+
"model.layers.28.mlp.gate_proj.weight": "model-00005-of-00009.safetensors",
|
200 |
+
"model.layers.28.mlp.up_proj.weight": "model-00005-of-00009.safetensors",
|
201 |
+
"model.layers.28.post_attention_layernorm.weight": "model-00005-of-00009.safetensors",
|
202 |
+
"model.layers.28.self_attn.k_proj.weight": "model-00005-of-00009.safetensors",
|
203 |
+
"model.layers.28.self_attn.o_proj.weight": "model-00005-of-00009.safetensors",
|
204 |
+
"model.layers.28.self_attn.q_proj.weight": "model-00005-of-00009.safetensors",
|
205 |
+
"model.layers.28.self_attn.v_proj.weight": "model-00005-of-00009.safetensors",
|
206 |
+
"model.layers.29.input_layernorm.weight": "model-00005-of-00009.safetensors",
|
207 |
+
"model.layers.29.mlp.down_proj.weight": "model-00005-of-00009.safetensors",
|
208 |
+
"model.layers.29.mlp.gate_proj.weight": "model-00005-of-00009.safetensors",
|
209 |
+
"model.layers.29.mlp.up_proj.weight": "model-00005-of-00009.safetensors",
|
210 |
+
"model.layers.29.post_attention_layernorm.weight": "model-00005-of-00009.safetensors",
|
211 |
+
"model.layers.29.self_attn.k_proj.weight": "model-00005-of-00009.safetensors",
|
212 |
+
"model.layers.29.self_attn.o_proj.weight": "model-00005-of-00009.safetensors",
|
213 |
+
"model.layers.29.self_attn.q_proj.weight": "model-00005-of-00009.safetensors",
|
214 |
+
"model.layers.29.self_attn.v_proj.weight": "model-00005-of-00009.safetensors",
|
215 |
+
"model.layers.3.input_layernorm.weight": "model-00001-of-00009.safetensors",
|
216 |
+
"model.layers.3.mlp.down_proj.weight": "model-00001-of-00009.safetensors",
|
217 |
+
"model.layers.3.mlp.gate_proj.weight": "model-00001-of-00009.safetensors",
|
218 |
+
"model.layers.3.mlp.up_proj.weight": "model-00001-of-00009.safetensors",
|
219 |
+
"model.layers.3.post_attention_layernorm.weight": "model-00001-of-00009.safetensors",
|
220 |
+
"model.layers.3.self_attn.k_proj.weight": "model-00001-of-00009.safetensors",
|
221 |
+
"model.layers.3.self_attn.o_proj.weight": "model-00001-of-00009.safetensors",
|
222 |
+
"model.layers.3.self_attn.q_proj.weight": "model-00001-of-00009.safetensors",
|
223 |
+
"model.layers.3.self_attn.v_proj.weight": "model-00001-of-00009.safetensors",
|
224 |
+
"model.layers.30.input_layernorm.weight": "model-00005-of-00009.safetensors",
|
225 |
+
"model.layers.30.mlp.down_proj.weight": "model-00005-of-00009.safetensors",
|
226 |
+
"model.layers.30.mlp.gate_proj.weight": "model-00005-of-00009.safetensors",
|
227 |
+
"model.layers.30.mlp.up_proj.weight": "model-00005-of-00009.safetensors",
|
228 |
+
"model.layers.30.post_attention_layernorm.weight": "model-00005-of-00009.safetensors",
|
229 |
+
"model.layers.30.self_attn.k_proj.weight": "model-00005-of-00009.safetensors",
|
230 |
+
"model.layers.30.self_attn.o_proj.weight": "model-00005-of-00009.safetensors",
|
231 |
+
"model.layers.30.self_attn.q_proj.weight": "model-00005-of-00009.safetensors",
|
232 |
+
"model.layers.30.self_attn.v_proj.weight": "model-00005-of-00009.safetensors",
|
233 |
+
"model.layers.31.input_layernorm.weight": "model-00006-of-00009.safetensors",
|
234 |
+
"model.layers.31.mlp.down_proj.weight": "model-00006-of-00009.safetensors",
|
235 |
+
"model.layers.31.mlp.gate_proj.weight": "model-00006-of-00009.safetensors",
|
236 |
+
"model.layers.31.mlp.up_proj.weight": "model-00006-of-00009.safetensors",
|
237 |
+
"model.layers.31.post_attention_layernorm.weight": "model-00006-of-00009.safetensors",
|
238 |
+
"model.layers.31.self_attn.k_proj.weight": "model-00005-of-00009.safetensors",
|
239 |
+
"model.layers.31.self_attn.o_proj.weight": "model-00006-of-00009.safetensors",
|
240 |
+
"model.layers.31.self_attn.q_proj.weight": "model-00005-of-00009.safetensors",
|
241 |
+
"model.layers.31.self_attn.v_proj.weight": "model-00005-of-00009.safetensors",
|
242 |
+
"model.layers.32.input_layernorm.weight": "model-00006-of-00009.safetensors",
|
243 |
+
"model.layers.32.mlp.down_proj.weight": "model-00006-of-00009.safetensors",
|
244 |
+
"model.layers.32.mlp.gate_proj.weight": "model-00006-of-00009.safetensors",
|
245 |
+
"model.layers.32.mlp.up_proj.weight": "model-00006-of-00009.safetensors",
|
246 |
+
"model.layers.32.post_attention_layernorm.weight": "model-00006-of-00009.safetensors",
|
247 |
+
"model.layers.32.self_attn.k_proj.weight": "model-00006-of-00009.safetensors",
|
248 |
+
"model.layers.32.self_attn.o_proj.weight": "model-00006-of-00009.safetensors",
|
249 |
+
"model.layers.32.self_attn.q_proj.weight": "model-00006-of-00009.safetensors",
|
250 |
+
"model.layers.32.self_attn.v_proj.weight": "model-00006-of-00009.safetensors",
|
251 |
+
"model.layers.33.input_layernorm.weight": "model-00006-of-00009.safetensors",
|
252 |
+
"model.layers.33.mlp.down_proj.weight": "model-00006-of-00009.safetensors",
|
253 |
+
"model.layers.33.mlp.gate_proj.weight": "model-00006-of-00009.safetensors",
|
254 |
+
"model.layers.33.mlp.up_proj.weight": "model-00006-of-00009.safetensors",
|
255 |
+
"model.layers.33.post_attention_layernorm.weight": "model-00006-of-00009.safetensors",
|
256 |
+
"model.layers.33.self_attn.k_proj.weight": "model-00006-of-00009.safetensors",
|
257 |
+
"model.layers.33.self_attn.o_proj.weight": "model-00006-of-00009.safetensors",
|
258 |
+
"model.layers.33.self_attn.q_proj.weight": "model-00006-of-00009.safetensors",
|
259 |
+
"model.layers.33.self_attn.v_proj.weight": "model-00006-of-00009.safetensors",
|
260 |
+
"model.layers.34.input_layernorm.weight": "model-00006-of-00009.safetensors",
|
261 |
+
"model.layers.34.mlp.down_proj.weight": "model-00006-of-00009.safetensors",
|
262 |
+
"model.layers.34.mlp.gate_proj.weight": "model-00006-of-00009.safetensors",
|
263 |
+
"model.layers.34.mlp.up_proj.weight": "model-00006-of-00009.safetensors",
|
264 |
+
"model.layers.34.post_attention_layernorm.weight": "model-00006-of-00009.safetensors",
|
265 |
+
"model.layers.34.self_attn.k_proj.weight": "model-00006-of-00009.safetensors",
|
266 |
+
"model.layers.34.self_attn.o_proj.weight": "model-00006-of-00009.safetensors",
|
267 |
+
"model.layers.34.self_attn.q_proj.weight": "model-00006-of-00009.safetensors",
|
268 |
+
"model.layers.34.self_attn.v_proj.weight": "model-00006-of-00009.safetensors",
|
269 |
+
"model.layers.35.input_layernorm.weight": "model-00006-of-00009.safetensors",
|
270 |
+
"model.layers.35.mlp.down_proj.weight": "model-00006-of-00009.safetensors",
|
271 |
+
"model.layers.35.mlp.gate_proj.weight": "model-00006-of-00009.safetensors",
|
272 |
+
"model.layers.35.mlp.up_proj.weight": "model-00006-of-00009.safetensors",
|
273 |
+
"model.layers.35.post_attention_layernorm.weight": "model-00006-of-00009.safetensors",
|
274 |
+
"model.layers.35.self_attn.k_proj.weight": "model-00006-of-00009.safetensors",
|
275 |
+
"model.layers.35.self_attn.o_proj.weight": "model-00006-of-00009.safetensors",
|
276 |
+
"model.layers.35.self_attn.q_proj.weight": "model-00006-of-00009.safetensors",
|
277 |
+
"model.layers.35.self_attn.v_proj.weight": "model-00006-of-00009.safetensors",
|
278 |
+
"model.layers.36.input_layernorm.weight": "model-00006-of-00009.safetensors",
|
279 |
+
"model.layers.36.mlp.down_proj.weight": "model-00006-of-00009.safetensors",
|
280 |
+
"model.layers.36.mlp.gate_proj.weight": "model-00006-of-00009.safetensors",
|
281 |
+
"model.layers.36.mlp.up_proj.weight": "model-00006-of-00009.safetensors",
|
282 |
+
"model.layers.36.post_attention_layernorm.weight": "model-00006-of-00009.safetensors",
|
283 |
+
"model.layers.36.self_attn.k_proj.weight": "model-00006-of-00009.safetensors",
|
284 |
+
"model.layers.36.self_attn.o_proj.weight": "model-00006-of-00009.safetensors",
|
285 |
+
"model.layers.36.self_attn.q_proj.weight": "model-00006-of-00009.safetensors",
|
286 |
+
"model.layers.36.self_attn.v_proj.weight": "model-00006-of-00009.safetensors",
|
287 |
+
"model.layers.37.input_layernorm.weight": "model-00007-of-00009.safetensors",
|
288 |
+
"model.layers.37.mlp.down_proj.weight": "model-00007-of-00009.safetensors",
|
289 |
+
"model.layers.37.mlp.gate_proj.weight": "model-00006-of-00009.safetensors",
|
290 |
+
"model.layers.37.mlp.up_proj.weight": "model-00007-of-00009.safetensors",
|
291 |
+
"model.layers.37.post_attention_layernorm.weight": "model-00007-of-00009.safetensors",
|
292 |
+
"model.layers.37.self_attn.k_proj.weight": "model-00006-of-00009.safetensors",
|
293 |
+
"model.layers.37.self_attn.o_proj.weight": "model-00006-of-00009.safetensors",
|
294 |
+
"model.layers.37.self_attn.q_proj.weight": "model-00006-of-00009.safetensors",
|
295 |
+
"model.layers.37.self_attn.v_proj.weight": "model-00006-of-00009.safetensors",
|
296 |
+
"model.layers.38.input_layernorm.weight": "model-00007-of-00009.safetensors",
|
297 |
+
"model.layers.38.mlp.down_proj.weight": "model-00007-of-00009.safetensors",
|
298 |
+
"model.layers.38.mlp.gate_proj.weight": "model-00007-of-00009.safetensors",
|
299 |
+
"model.layers.38.mlp.up_proj.weight": "model-00007-of-00009.safetensors",
|
300 |
+
"model.layers.38.post_attention_layernorm.weight": "model-00007-of-00009.safetensors",
|
301 |
+
"model.layers.38.self_attn.k_proj.weight": "model-00007-of-00009.safetensors",
|
302 |
+
"model.layers.38.self_attn.o_proj.weight": "model-00007-of-00009.safetensors",
|
303 |
+
"model.layers.38.self_attn.q_proj.weight": "model-00007-of-00009.safetensors",
|
304 |
+
"model.layers.38.self_attn.v_proj.weight": "model-00007-of-00009.safetensors",
|
305 |
+
"model.layers.39.input_layernorm.weight": "model-00007-of-00009.safetensors",
|
306 |
+
"model.layers.39.mlp.down_proj.weight": "model-00007-of-00009.safetensors",
|
307 |
+
"model.layers.39.mlp.gate_proj.weight": "model-00007-of-00009.safetensors",
|
308 |
+
"model.layers.39.mlp.up_proj.weight": "model-00007-of-00009.safetensors",
|
309 |
+
"model.layers.39.post_attention_layernorm.weight": "model-00007-of-00009.safetensors",
|
310 |
+
"model.layers.39.self_attn.k_proj.weight": "model-00007-of-00009.safetensors",
|
311 |
+
"model.layers.39.self_attn.o_proj.weight": "model-00007-of-00009.safetensors",
|
312 |
+
"model.layers.39.self_attn.q_proj.weight": "model-00007-of-00009.safetensors",
|
313 |
+
"model.layers.39.self_attn.v_proj.weight": "model-00007-of-00009.safetensors",
|
314 |
+
"model.layers.4.input_layernorm.weight": "model-00001-of-00009.safetensors",
|
315 |
+
"model.layers.4.mlp.down_proj.weight": "model-00001-of-00009.safetensors",
|
316 |
+
"model.layers.4.mlp.gate_proj.weight": "model-00001-of-00009.safetensors",
|
317 |
+
"model.layers.4.mlp.up_proj.weight": "model-00001-of-00009.safetensors",
|
318 |
+
"model.layers.4.post_attention_layernorm.weight": "model-00001-of-00009.safetensors",
|
319 |
+
"model.layers.4.self_attn.k_proj.weight": "model-00001-of-00009.safetensors",
|
320 |
+
"model.layers.4.self_attn.o_proj.weight": "model-00001-of-00009.safetensors",
|
321 |
+
"model.layers.4.self_attn.q_proj.weight": "model-00001-of-00009.safetensors",
|
322 |
+
"model.layers.4.self_attn.v_proj.weight": "model-00001-of-00009.safetensors",
|
323 |
+
"model.layers.40.input_layernorm.weight": "model-00007-of-00009.safetensors",
|
324 |
+
"model.layers.40.mlp.down_proj.weight": "model-00007-of-00009.safetensors",
|
325 |
+
"model.layers.40.mlp.gate_proj.weight": "model-00007-of-00009.safetensors",
|
326 |
+
"model.layers.40.mlp.up_proj.weight": "model-00007-of-00009.safetensors",
|
327 |
+
"model.layers.40.post_attention_layernorm.weight": "model-00007-of-00009.safetensors",
|
328 |
+
"model.layers.40.self_attn.k_proj.weight": "model-00007-of-00009.safetensors",
|
329 |
+
"model.layers.40.self_attn.o_proj.weight": "model-00007-of-00009.safetensors",
|
330 |
+
"model.layers.40.self_attn.q_proj.weight": "model-00007-of-00009.safetensors",
|
331 |
+
"model.layers.40.self_attn.v_proj.weight": "model-00007-of-00009.safetensors",
|
332 |
+
"model.layers.41.input_layernorm.weight": "model-00007-of-00009.safetensors",
|
333 |
+
"model.layers.41.mlp.down_proj.weight": "model-00007-of-00009.safetensors",
|
334 |
+
"model.layers.41.mlp.gate_proj.weight": "model-00007-of-00009.safetensors",
|
335 |
+
"model.layers.41.mlp.up_proj.weight": "model-00007-of-00009.safetensors",
|
336 |
+
"model.layers.41.post_attention_layernorm.weight": "model-00007-of-00009.safetensors",
|
337 |
+
"model.layers.41.self_attn.k_proj.weight": "model-00007-of-00009.safetensors",
|
338 |
+
"model.layers.41.self_attn.o_proj.weight": "model-00007-of-00009.safetensors",
|
339 |
+
"model.layers.41.self_attn.q_proj.weight": "model-00007-of-00009.safetensors",
|
340 |
+
"model.layers.41.self_attn.v_proj.weight": "model-00007-of-00009.safetensors",
|
341 |
+
"model.layers.42.input_layernorm.weight": "model-00007-of-00009.safetensors",
|
342 |
+
"model.layers.42.mlp.down_proj.weight": "model-00007-of-00009.safetensors",
|
343 |
+
"model.layers.42.mlp.gate_proj.weight": "model-00007-of-00009.safetensors",
|
344 |
+
"model.layers.42.mlp.up_proj.weight": "model-00007-of-00009.safetensors",
|
345 |
+
"model.layers.42.post_attention_layernorm.weight": "model-00007-of-00009.safetensors",
|
346 |
+
"model.layers.42.self_attn.k_proj.weight": "model-00007-of-00009.safetensors",
|
347 |
+
"model.layers.42.self_attn.o_proj.weight": "model-00007-of-00009.safetensors",
|
348 |
+
"model.layers.42.self_attn.q_proj.weight": "model-00007-of-00009.safetensors",
|
349 |
+
"model.layers.42.self_attn.v_proj.weight": "model-00007-of-00009.safetensors",
|
350 |
+
"model.layers.43.input_layernorm.weight": "model-00008-of-00009.safetensors",
|
351 |
+
"model.layers.43.mlp.down_proj.weight": "model-00008-of-00009.safetensors",
|
352 |
+
"model.layers.43.mlp.gate_proj.weight": "model-00007-of-00009.safetensors",
|
353 |
+
"model.layers.43.mlp.up_proj.weight": "model-00007-of-00009.safetensors",
|
354 |
+
"model.layers.43.post_attention_layernorm.weight": "model-00008-of-00009.safetensors",
|
355 |
+
"model.layers.43.self_attn.k_proj.weight": "model-00007-of-00009.safetensors",
|
356 |
+
"model.layers.43.self_attn.o_proj.weight": "model-00007-of-00009.safetensors",
|
357 |
+
"model.layers.43.self_attn.q_proj.weight": "model-00007-of-00009.safetensors",
|
358 |
+
"model.layers.43.self_attn.v_proj.weight": "model-00007-of-00009.safetensors",
|
359 |
+
"model.layers.44.input_layernorm.weight": "model-00008-of-00009.safetensors",
|
360 |
+
"model.layers.44.mlp.down_proj.weight": "model-00008-of-00009.safetensors",
|
361 |
+
"model.layers.44.mlp.gate_proj.weight": "model-00008-of-00009.safetensors",
|
362 |
+
"model.layers.44.mlp.up_proj.weight": "model-00008-of-00009.safetensors",
|
363 |
+
"model.layers.44.post_attention_layernorm.weight": "model-00008-of-00009.safetensors",
|
364 |
+
"model.layers.44.self_attn.k_proj.weight": "model-00008-of-00009.safetensors",
|
365 |
+
"model.layers.44.self_attn.o_proj.weight": "model-00008-of-00009.safetensors",
|
366 |
+
"model.layers.44.self_attn.q_proj.weight": "model-00008-of-00009.safetensors",
|
367 |
+
"model.layers.44.self_attn.v_proj.weight": "model-00008-of-00009.safetensors",
|
368 |
+
"model.layers.45.input_layernorm.weight": "model-00008-of-00009.safetensors",
|
369 |
+
"model.layers.45.mlp.down_proj.weight": "model-00008-of-00009.safetensors",
|
370 |
+
"model.layers.45.mlp.gate_proj.weight": "model-00008-of-00009.safetensors",
|
371 |
+
"model.layers.45.mlp.up_proj.weight": "model-00008-of-00009.safetensors",
|
372 |
+
"model.layers.45.post_attention_layernorm.weight": "model-00008-of-00009.safetensors",
|
373 |
+
"model.layers.45.self_attn.k_proj.weight": "model-00008-of-00009.safetensors",
|
374 |
+
"model.layers.45.self_attn.o_proj.weight": "model-00008-of-00009.safetensors",
|
375 |
+
"model.layers.45.self_attn.q_proj.weight": "model-00008-of-00009.safetensors",
|
376 |
+
"model.layers.45.self_attn.v_proj.weight": "model-00008-of-00009.safetensors",
|
377 |
+
"model.layers.46.input_layernorm.weight": "model-00008-of-00009.safetensors",
|
378 |
+
"model.layers.46.mlp.down_proj.weight": "model-00008-of-00009.safetensors",
|
379 |
+
"model.layers.46.mlp.gate_proj.weight": "model-00008-of-00009.safetensors",
|
380 |
+
"model.layers.46.mlp.up_proj.weight": "model-00008-of-00009.safetensors",
|
381 |
+
"model.layers.46.post_attention_layernorm.weight": "model-00008-of-00009.safetensors",
|
382 |
+
"model.layers.46.self_attn.k_proj.weight": "model-00008-of-00009.safetensors",
|
383 |
+
"model.layers.46.self_attn.o_proj.weight": "model-00008-of-00009.safetensors",
|
384 |
+
"model.layers.46.self_attn.q_proj.weight": "model-00008-of-00009.safetensors",
|
385 |
+
"model.layers.46.self_attn.v_proj.weight": "model-00008-of-00009.safetensors",
|
386 |
+
"model.layers.47.input_layernorm.weight": "model-00008-of-00009.safetensors",
|
387 |
+
"model.layers.47.mlp.down_proj.weight": "model-00008-of-00009.safetensors",
|
388 |
+
"model.layers.47.mlp.gate_proj.weight": "model-00008-of-00009.safetensors",
|
389 |
+
"model.layers.47.mlp.up_proj.weight": "model-00008-of-00009.safetensors",
|
390 |
+
"model.layers.47.post_attention_layernorm.weight": "model-00008-of-00009.safetensors",
|
391 |
+
"model.layers.47.self_attn.k_proj.weight": "model-00008-of-00009.safetensors",
|
392 |
+
"model.layers.47.self_attn.o_proj.weight": "model-00008-of-00009.safetensors",
|
393 |
+
"model.layers.47.self_attn.q_proj.weight": "model-00008-of-00009.safetensors",
|
394 |
+
"model.layers.47.self_attn.v_proj.weight": "model-00008-of-00009.safetensors",
|
395 |
+
"model.layers.48.input_layernorm.weight": "model-00008-of-00009.safetensors",
|
396 |
+
"model.layers.48.mlp.down_proj.weight": "model-00008-of-00009.safetensors",
|
397 |
+
"model.layers.48.mlp.gate_proj.weight": "model-00008-of-00009.safetensors",
|
398 |
+
"model.layers.48.mlp.up_proj.weight": "model-00008-of-00009.safetensors",
|
399 |
+
"model.layers.48.post_attention_layernorm.weight": "model-00008-of-00009.safetensors",
|
400 |
+
"model.layers.48.self_attn.k_proj.weight": "model-00008-of-00009.safetensors",
|
401 |
+
"model.layers.48.self_attn.o_proj.weight": "model-00008-of-00009.safetensors",
|
402 |
+
"model.layers.48.self_attn.q_proj.weight": "model-00008-of-00009.safetensors",
|
403 |
+
"model.layers.48.self_attn.v_proj.weight": "model-00008-of-00009.safetensors",
|
404 |
+
"model.layers.49.input_layernorm.weight": "model-00008-of-00009.safetensors",
|
405 |
+
"model.layers.49.mlp.down_proj.weight": "model-00008-of-00009.safetensors",
|
406 |
+
"model.layers.49.mlp.gate_proj.weight": "model-00008-of-00009.safetensors",
|
407 |
+
"model.layers.49.mlp.up_proj.weight": "model-00008-of-00009.safetensors",
|
408 |
+
"model.layers.49.post_attention_layernorm.weight": "model-00008-of-00009.safetensors",
|
409 |
+
"model.layers.49.self_attn.k_proj.weight": "model-00008-of-00009.safetensors",
|
410 |
+
"model.layers.49.self_attn.o_proj.weight": "model-00008-of-00009.safetensors",
|
411 |
+
"model.layers.49.self_attn.q_proj.weight": "model-00008-of-00009.safetensors",
|
412 |
+
"model.layers.49.self_attn.v_proj.weight": "model-00008-of-00009.safetensors",
|
413 |
+
"model.layers.5.input_layernorm.weight": "model-00002-of-00009.safetensors",
|
414 |
+
"model.layers.5.mlp.down_proj.weight": "model-00002-of-00009.safetensors",
|
415 |
+
"model.layers.5.mlp.gate_proj.weight": "model-00001-of-00009.safetensors",
|
416 |
+
"model.layers.5.mlp.up_proj.weight": "model-00001-of-00009.safetensors",
|
417 |
+
"model.layers.5.post_attention_layernorm.weight": "model-00002-of-00009.safetensors",
|
418 |
+
"model.layers.5.self_attn.k_proj.weight": "model-00001-of-00009.safetensors",
|
419 |
+
"model.layers.5.self_attn.o_proj.weight": "model-00001-of-00009.safetensors",
|
420 |
+
"model.layers.5.self_attn.q_proj.weight": "model-00001-of-00009.safetensors",
|
421 |
+
"model.layers.5.self_attn.v_proj.weight": "model-00001-of-00009.safetensors",
|
422 |
+
"model.layers.50.input_layernorm.weight": "model-00009-of-00009.safetensors",
|
423 |
+
"model.layers.50.mlp.down_proj.weight": "model-00009-of-00009.safetensors",
|
424 |
+
"model.layers.50.mlp.gate_proj.weight": "model-00009-of-00009.safetensors",
|
425 |
+
"model.layers.50.mlp.up_proj.weight": "model-00009-of-00009.safetensors",
|
426 |
+
"model.layers.50.post_attention_layernorm.weight": "model-00009-of-00009.safetensors",
|
427 |
+
"model.layers.50.self_attn.k_proj.weight": "model-00008-of-00009.safetensors",
|
428 |
+
"model.layers.50.self_attn.o_proj.weight": "model-00009-of-00009.safetensors",
|
429 |
+
"model.layers.50.self_attn.q_proj.weight": "model-00008-of-00009.safetensors",
|
430 |
+
"model.layers.50.self_attn.v_proj.weight": "model-00008-of-00009.safetensors",
|
431 |
+
"model.layers.51.input_layernorm.weight": "model-00009-of-00009.safetensors",
|
432 |
+
"model.layers.51.mlp.down_proj.weight": "model-00009-of-00009.safetensors",
|
433 |
+
"model.layers.51.mlp.gate_proj.weight": "model-00009-of-00009.safetensors",
|
434 |
+
"model.layers.51.mlp.up_proj.weight": "model-00009-of-00009.safetensors",
|
435 |
+
"model.layers.51.post_attention_layernorm.weight": "model-00009-of-00009.safetensors",
|
436 |
+
"model.layers.51.self_attn.k_proj.weight": "model-00009-of-00009.safetensors",
|
437 |
+
"model.layers.51.self_attn.o_proj.weight": "model-00009-of-00009.safetensors",
|
438 |
+
"model.layers.51.self_attn.q_proj.weight": "model-00009-of-00009.safetensors",
|
439 |
+
"model.layers.51.self_attn.v_proj.weight": "model-00009-of-00009.safetensors",
|
440 |
+
"model.layers.52.input_layernorm.weight": "model-00009-of-00009.safetensors",
|
441 |
+
"model.layers.52.mlp.down_proj.weight": "model-00009-of-00009.safetensors",
|
442 |
+
"model.layers.52.mlp.gate_proj.weight": "model-00009-of-00009.safetensors",
|
443 |
+
"model.layers.52.mlp.up_proj.weight": "model-00009-of-00009.safetensors",
|
444 |
+
"model.layers.52.post_attention_layernorm.weight": "model-00009-of-00009.safetensors",
|
445 |
+
"model.layers.52.self_attn.k_proj.weight": "model-00009-of-00009.safetensors",
|
446 |
+
"model.layers.52.self_attn.o_proj.weight": "model-00009-of-00009.safetensors",
|
447 |
+
"model.layers.52.self_attn.q_proj.weight": "model-00009-of-00009.safetensors",
|
448 |
+
"model.layers.52.self_attn.v_proj.weight": "model-00009-of-00009.safetensors",
|
449 |
+
"model.layers.53.input_layernorm.weight": "model-00009-of-00009.safetensors",
|
450 |
+
"model.layers.53.mlp.down_proj.weight": "model-00009-of-00009.safetensors",
|
451 |
+
"model.layers.53.mlp.gate_proj.weight": "model-00009-of-00009.safetensors",
|
452 |
+
"model.layers.53.mlp.up_proj.weight": "model-00009-of-00009.safetensors",
|
453 |
+
"model.layers.53.post_attention_layernorm.weight": "model-00009-of-00009.safetensors",
|
454 |
+
"model.layers.53.self_attn.k_proj.weight": "model-00009-of-00009.safetensors",
|
455 |
+
"model.layers.53.self_attn.o_proj.weight": "model-00009-of-00009.safetensors",
|
456 |
+
"model.layers.53.self_attn.q_proj.weight": "model-00009-of-00009.safetensors",
|
457 |
+
"model.layers.53.self_attn.v_proj.weight": "model-00009-of-00009.safetensors",
|
458 |
+
"model.layers.54.input_layernorm.weight": "model-00009-of-00009.safetensors",
|
459 |
+
"model.layers.54.mlp.down_proj.weight": "model-00009-of-00009.safetensors",
|
460 |
+
"model.layers.54.mlp.gate_proj.weight": "model-00009-of-00009.safetensors",
|
461 |
+
"model.layers.54.mlp.up_proj.weight": "model-00009-of-00009.safetensors",
|
462 |
+
"model.layers.54.post_attention_layernorm.weight": "model-00009-of-00009.safetensors",
|
463 |
+
"model.layers.54.self_attn.k_proj.weight": "model-00009-of-00009.safetensors",
|
464 |
+
"model.layers.54.self_attn.o_proj.weight": "model-00009-of-00009.safetensors",
|
465 |
+
"model.layers.54.self_attn.q_proj.weight": "model-00009-of-00009.safetensors",
|
466 |
+
"model.layers.54.self_attn.v_proj.weight": "model-00009-of-00009.safetensors",
|
467 |
+
"model.layers.55.input_layernorm.weight": "model-00009-of-00009.safetensors",
|
468 |
+
"model.layers.55.mlp.down_proj.weight": "model-00009-of-00009.safetensors",
|
469 |
+
"model.layers.55.mlp.gate_proj.weight": "model-00009-of-00009.safetensors",
|
470 |
+
"model.layers.55.mlp.up_proj.weight": "model-00009-of-00009.safetensors",
|
471 |
+
"model.layers.55.post_attention_layernorm.weight": "model-00009-of-00009.safetensors",
|
472 |
+
"model.layers.55.self_attn.k_proj.weight": "model-00009-of-00009.safetensors",
|
473 |
+
"model.layers.55.self_attn.o_proj.weight": "model-00009-of-00009.safetensors",
|
474 |
+
"model.layers.55.self_attn.q_proj.weight": "model-00009-of-00009.safetensors",
|
475 |
+
"model.layers.55.self_attn.v_proj.weight": "model-00009-of-00009.safetensors",
|
476 |
+
"model.layers.6.input_layernorm.weight": "model-00002-of-00009.safetensors",
|
477 |
+
"model.layers.6.mlp.down_proj.weight": "model-00002-of-00009.safetensors",
|
478 |
+
"model.layers.6.mlp.gate_proj.weight": "model-00002-of-00009.safetensors",
|
479 |
+
"model.layers.6.mlp.up_proj.weight": "model-00002-of-00009.safetensors",
|
480 |
+
"model.layers.6.post_attention_layernorm.weight": "model-00002-of-00009.safetensors",
|
481 |
+
"model.layers.6.self_attn.k_proj.weight": "model-00002-of-00009.safetensors",
|
482 |
+
"model.layers.6.self_attn.o_proj.weight": "model-00002-of-00009.safetensors",
|
483 |
+
"model.layers.6.self_attn.q_proj.weight": "model-00002-of-00009.safetensors",
|
484 |
+
"model.layers.6.self_attn.v_proj.weight": "model-00002-of-00009.safetensors",
|
485 |
+
"model.layers.7.input_layernorm.weight": "model-00002-of-00009.safetensors",
|
486 |
+
"model.layers.7.mlp.down_proj.weight": "model-00002-of-00009.safetensors",
|
487 |
+
"model.layers.7.mlp.gate_proj.weight": "model-00002-of-00009.safetensors",
|
488 |
+
"model.layers.7.mlp.up_proj.weight": "model-00002-of-00009.safetensors",
|
489 |
+
"model.layers.7.post_attention_layernorm.weight": "model-00002-of-00009.safetensors",
|
490 |
+
"model.layers.7.self_attn.k_proj.weight": "model-00002-of-00009.safetensors",
|
491 |
+
"model.layers.7.self_attn.o_proj.weight": "model-00002-of-00009.safetensors",
|
492 |
+
"model.layers.7.self_attn.q_proj.weight": "model-00002-of-00009.safetensors",
|
493 |
+
"model.layers.7.self_attn.v_proj.weight": "model-00002-of-00009.safetensors",
|
494 |
+
"model.layers.8.input_layernorm.weight": "model-00002-of-00009.safetensors",
|
495 |
+
"model.layers.8.mlp.down_proj.weight": "model-00002-of-00009.safetensors",
|
496 |
+
"model.layers.8.mlp.gate_proj.weight": "model-00002-of-00009.safetensors",
|
497 |
+
"model.layers.8.mlp.up_proj.weight": "model-00002-of-00009.safetensors",
|
498 |
+
"model.layers.8.post_attention_layernorm.weight": "model-00002-of-00009.safetensors",
|
499 |
+
"model.layers.8.self_attn.k_proj.weight": "model-00002-of-00009.safetensors",
|
500 |
+
"model.layers.8.self_attn.o_proj.weight": "model-00002-of-00009.safetensors",
|
501 |
+
"model.layers.8.self_attn.q_proj.weight": "model-00002-of-00009.safetensors",
|
502 |
+
"model.layers.8.self_attn.v_proj.weight": "model-00002-of-00009.safetensors",
|
503 |
+
"model.layers.9.input_layernorm.weight": "model-00002-of-00009.safetensors",
|
504 |
+
"model.layers.9.mlp.down_proj.weight": "model-00002-of-00009.safetensors",
|
505 |
+
"model.layers.9.mlp.gate_proj.weight": "model-00002-of-00009.safetensors",
|
506 |
+
"model.layers.9.mlp.up_proj.weight": "model-00002-of-00009.safetensors",
|
507 |
+
"model.layers.9.post_attention_layernorm.weight": "model-00002-of-00009.safetensors",
|
508 |
+
"model.layers.9.self_attn.k_proj.weight": "model-00002-of-00009.safetensors",
|
509 |
+
"model.layers.9.self_attn.o_proj.weight": "model-00002-of-00009.safetensors",
|
510 |
+
"model.layers.9.self_attn.q_proj.weight": "model-00002-of-00009.safetensors",
|
511 |
+
"model.layers.9.self_attn.v_proj.weight": "model-00002-of-00009.safetensors",
|
512 |
+
"model.norm.weight": "model-00009-of-00009.safetensors"
|
513 |
+
}
|
514 |
+
}
|
output-00001-of-00002.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:91daf2e31e7ccbdf576a1e3604782db712516653ef15e1a6e4637faf0fc69a5b
|
3 |
+
size 8574566900
|
output-00002-of-00002.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:53cc1fa83037c84ef6c0a31407ffa4b92c4006340a7c0391c61f734a8bd40ef3
|
3 |
+
size 1586297648
|
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 sha256:37f00374dea48658ee8f5d0f21895b9bc55cb0103939607c8185bfd1c6ca1f89
|
3 |
+
size 587404
|
tokenizer_config.json
ADDED
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"add_bos_token": true,
|
3 |
+
"add_eos_token": false,
|
4 |
+
"add_prefix_space": true,
|
5 |
+
"added_tokens_decoder": {
|
6 |
+
"0": {
|
7 |
+
"content": "<unk>",
|
8 |
+
"lstrip": false,
|
9 |
+
"normalized": false,
|
10 |
+
"rstrip": false,
|
11 |
+
"single_word": false,
|
12 |
+
"special": true
|
13 |
+
},
|
14 |
+
"1": {
|
15 |
+
"content": "<s>",
|
16 |
+
"lstrip": false,
|
17 |
+
"normalized": false,
|
18 |
+
"rstrip": false,
|
19 |
+
"single_word": false,
|
20 |
+
"special": true
|
21 |
+
},
|
22 |
+
"2": {
|
23 |
+
"content": "</s>",
|
24 |
+
"lstrip": false,
|
25 |
+
"normalized": false,
|
26 |
+
"rstrip": false,
|
27 |
+
"single_word": false,
|
28 |
+
"special": true
|
29 |
+
}
|
30 |
+
},
|
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 |
+
}
|