RichardErkhov
commited on
Commit
•
2573f94
1
Parent(s):
59eedd4
uploaded readme
Browse files
README.md
ADDED
@@ -0,0 +1,199 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
Quantization made by Richard Erkhov.
|
2 |
+
|
3 |
+
[Github](https://github.com/RichardErkhov)
|
4 |
+
|
5 |
+
[Discord](https://discord.gg/pvy7H8DZMG)
|
6 |
+
|
7 |
+
[Request more models](https://github.com/RichardErkhov/quant_request)
|
8 |
+
|
9 |
+
|
10 |
+
sparse_mistral_7b_refined_web_50p_2024-04-13 - GGUF
|
11 |
+
- Model creator: https://huggingface.co/thrunlab/
|
12 |
+
- Original model: https://huggingface.co/thrunlab/sparse_mistral_7b_refined_web_50p_2024-04-13/
|
13 |
+
|
14 |
+
|
15 |
+
| Name | Quant method | Size |
|
16 |
+
| ---- | ---- | ---- |
|
17 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q2_K.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q2_K.gguf) | Q2_K | 2.53GB |
|
18 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.IQ3_XS.gguf) | IQ3_XS | 2.81GB |
|
19 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.IQ3_S.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.IQ3_S.gguf) | IQ3_S | 2.96GB |
|
20 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q3_K_S.gguf) | Q3_K_S | 2.95GB |
|
21 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.IQ3_M.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.IQ3_M.gguf) | IQ3_M | 3.06GB |
|
22 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q3_K.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q3_K.gguf) | Q3_K | 3.28GB |
|
23 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q3_K_M.gguf) | Q3_K_M | 3.28GB |
|
24 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q3_K_L.gguf) | Q3_K_L | 3.56GB |
|
25 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.IQ4_XS.gguf) | IQ4_XS | 3.67GB |
|
26 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q4_0.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q4_0.gguf) | Q4_0 | 3.83GB |
|
27 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.IQ4_NL.gguf) | IQ4_NL | 3.87GB |
|
28 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q4_K_S.gguf) | Q4_K_S | 3.86GB |
|
29 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q4_K.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q4_K.gguf) | Q4_K | 4.07GB |
|
30 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q4_K_M.gguf) | Q4_K_M | 4.07GB |
|
31 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q4_1.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q4_1.gguf) | Q4_1 | 4.24GB |
|
32 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q5_0.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q5_0.gguf) | Q5_0 | 4.65GB |
|
33 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q5_K_S.gguf) | Q5_K_S | 4.65GB |
|
34 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q5_K.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q5_K.gguf) | Q5_K | 4.78GB |
|
35 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q5_K_M.gguf) | Q5_K_M | 4.78GB |
|
36 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q5_1.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q5_1.gguf) | Q5_1 | 5.07GB |
|
37 |
+
| [sparse_mistral_7b_refined_web_50p_2024-04-13.Q6_K.gguf](https://huggingface.co/RichardErkhov/thrunlab_-_sparse_mistral_7b_refined_web_50p_2024-04-13-gguf/blob/main/sparse_mistral_7b_refined_web_50p_2024-04-13.Q6_K.gguf) | Q6_K | 5.53GB |
|
38 |
+
|
39 |
+
|
40 |
+
|
41 |
+
|
42 |
+
Original model description:
|
43 |
+
---
|
44 |
+
license: apache-2.0
|
45 |
+
base_model: mistralai/Mistral-7B-v0.1
|
46 |
+
tags:
|
47 |
+
- generated_from_trainer
|
48 |
+
model-index:
|
49 |
+
- name: sparse_mistral_7b_refined_web_50p_2024-04-13
|
50 |
+
results: []
|
51 |
+
---
|
52 |
+
|
53 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
54 |
+
should probably proofread and complete it, then remove this comment. -->
|
55 |
+
|
56 |
+
# sparse_mistral_7b_refined_web_50p_2024-04-13
|
57 |
+
|
58 |
+
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the None dataset.
|
59 |
+
It achieves the following results on the evaluation set:
|
60 |
+
- Loss: 2.1985
|
61 |
+
|
62 |
+
## Model description
|
63 |
+
|
64 |
+
More information needed
|
65 |
+
|
66 |
+
## Intended uses & limitations
|
67 |
+
|
68 |
+
More information needed
|
69 |
+
|
70 |
+
## Training and evaluation data
|
71 |
+
|
72 |
+
More information needed
|
73 |
+
|
74 |
+
## Training procedure
|
75 |
+
|
76 |
+
### Training hyperparameters
|
77 |
+
|
78 |
+
The following hyperparameters were used during training:
|
79 |
+
- learning_rate: 1e-05
|
80 |
+
- train_batch_size: 1
|
81 |
+
- eval_batch_size: 4
|
82 |
+
- seed: 0
|
83 |
+
- distributed_type: multi-GPU
|
84 |
+
- num_devices: 4
|
85 |
+
- gradient_accumulation_steps: 8
|
86 |
+
- total_train_batch_size: 32
|
87 |
+
- total_eval_batch_size: 16
|
88 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
89 |
+
- lr_scheduler_type: linear
|
90 |
+
- training_steps: 2350
|
91 |
+
|
92 |
+
### Training results
|
93 |
+
|
94 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
95 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
96 |
+
| 2.3391 | 0.01 | 25 | 2.4196 |
|
97 |
+
| 2.2711 | 0.02 | 50 | 2.3577 |
|
98 |
+
| 2.3054 | 0.02 | 75 | 2.3158 |
|
99 |
+
| 2.2795 | 0.03 | 100 | 2.2966 |
|
100 |
+
| 2.3175 | 0.04 | 125 | 2.2846 |
|
101 |
+
| 2.2388 | 0.05 | 150 | 2.2766 |
|
102 |
+
| 2.1679 | 0.06 | 175 | 2.2705 |
|
103 |
+
| 2.2996 | 0.06 | 200 | 2.2678 |
|
104 |
+
| 2.2788 | 0.07 | 225 | 2.2647 |
|
105 |
+
| 2.2448 | 0.08 | 250 | 2.2637 |
|
106 |
+
| 2.1837 | 0.09 | 275 | 2.2624 |
|
107 |
+
| 2.2089 | 0.1 | 300 | 2.2621 |
|
108 |
+
| 2.2686 | 0.1 | 325 | 2.2601 |
|
109 |
+
| 2.2254 | 0.11 | 350 | 2.2593 |
|
110 |
+
| 2.162 | 0.12 | 375 | 2.2590 |
|
111 |
+
| 2.2687 | 0.13 | 400 | 2.2563 |
|
112 |
+
| 2.2595 | 0.14 | 425 | 2.2571 |
|
113 |
+
| 2.186 | 0.14 | 450 | 2.2564 |
|
114 |
+
| 2.2689 | 0.15 | 475 | 2.2580 |
|
115 |
+
| 2.2472 | 0.16 | 500 | 2.2554 |
|
116 |
+
| 2.2005 | 0.17 | 525 | 2.2553 |
|
117 |
+
| 2.1983 | 0.18 | 550 | 2.2552 |
|
118 |
+
| 2.2388 | 0.18 | 575 | 2.2547 |
|
119 |
+
| 2.1443 | 0.19 | 600 | 2.2555 |
|
120 |
+
| 2.2198 | 0.2 | 625 | 2.2534 |
|
121 |
+
| 2.3008 | 0.21 | 650 | 2.2536 |
|
122 |
+
| 2.179 | 0.22 | 675 | 2.2521 |
|
123 |
+
| 2.2069 | 0.22 | 700 | 2.2531 |
|
124 |
+
| 2.1819 | 0.23 | 725 | 2.2526 |
|
125 |
+
| 2.1218 | 0.24 | 750 | 2.2536 |
|
126 |
+
| 2.1845 | 0.25 | 775 | 2.2515 |
|
127 |
+
| 2.2167 | 0.26 | 800 | 2.2510 |
|
128 |
+
| 2.2252 | 0.26 | 825 | 2.2520 |
|
129 |
+
| 2.1664 | 0.27 | 850 | 2.2519 |
|
130 |
+
| 2.1853 | 0.28 | 875 | 2.2530 |
|
131 |
+
| 2.1499 | 0.29 | 900 | 2.2513 |
|
132 |
+
| 2.2763 | 0.3 | 925 | 2.2517 |
|
133 |
+
| 2.2528 | 0.3 | 950 | 2.2518 |
|
134 |
+
| 2.2505 | 0.31 | 975 | 2.2500 |
|
135 |
+
| 2.1683 | 0.32 | 1000 | 2.2502 |
|
136 |
+
| 2.2177 | 0.33 | 1025 | 2.2501 |
|
137 |
+
| 2.238 | 0.34 | 1050 | 2.2516 |
|
138 |
+
| 2.193 | 0.34 | 1075 | 2.2507 |
|
139 |
+
| 2.2025 | 0.35 | 1100 | 2.2502 |
|
140 |
+
| 2.0944 | 0.36 | 1125 | 2.2512 |
|
141 |
+
| 2.2272 | 0.37 | 1150 | 2.2508 |
|
142 |
+
| 2.2264 | 0.38 | 1175 | 2.2500 |
|
143 |
+
| 2.1837 | 0.38 | 1200 | 2.2507 |
|
144 |
+
| 2.1444 | 0.39 | 1225 | 2.2489 |
|
145 |
+
| 2.2464 | 0.4 | 1250 | 2.2499 |
|
146 |
+
| 2.1388 | 0.41 | 1275 | 2.2508 |
|
147 |
+
| 2.193 | 0.42 | 1300 | 2.2492 |
|
148 |
+
| 2.2376 | 0.42 | 1325 | 2.2506 |
|
149 |
+
| 2.2212 | 0.43 | 1350 | 2.2478 |
|
150 |
+
| 2.2002 | 0.44 | 1375 | 2.2488 |
|
151 |
+
| 2.2729 | 0.45 | 1400 | 2.2484 |
|
152 |
+
| 2.2329 | 0.46 | 1425 | 2.2473 |
|
153 |
+
| 2.1919 | 0.46 | 1450 | 2.2481 |
|
154 |
+
| 2.2102 | 0.47 | 1475 | 2.2475 |
|
155 |
+
| 2.1466 | 0.48 | 1500 | 2.2473 |
|
156 |
+
| 2.1819 | 0.49 | 1525 | 2.2478 |
|
157 |
+
| 2.2558 | 0.5 | 1550 | 2.2468 |
|
158 |
+
| 2.2137 | 0.5 | 1575 | 2.2463 |
|
159 |
+
| 2.2288 | 0.51 | 1600 | 2.2466 |
|
160 |
+
| 2.1479 | 0.52 | 1625 | 2.2468 |
|
161 |
+
| 2.1726 | 0.53 | 1650 | 2.2471 |
|
162 |
+
| 2.1805 | 0.54 | 1675 | 2.2454 |
|
163 |
+
| 2.1505 | 0.54 | 1700 | 2.2470 |
|
164 |
+
| 2.1337 | 0.55 | 1725 | 2.2465 |
|
165 |
+
| 2.2413 | 0.56 | 1750 | 2.2460 |
|
166 |
+
| 2.152 | 0.57 | 1775 | 2.2478 |
|
167 |
+
| 2.2669 | 0.58 | 1800 | 2.2471 |
|
168 |
+
| 2.2925 | 0.58 | 1825 | 2.2465 |
|
169 |
+
| 2.222 | 0.59 | 1850 | 2.2457 |
|
170 |
+
| 2.1308 | 0.6 | 1875 | 2.2466 |
|
171 |
+
| 2.201 | 0.61 | 1900 | 2.2456 |
|
172 |
+
| 2.2247 | 0.62 | 1925 | 2.2460 |
|
173 |
+
| 2.2426 | 0.62 | 1950 | 2.2463 |
|
174 |
+
| 2.2312 | 0.63 | 1975 | 2.2465 |
|
175 |
+
| 2.2679 | 0.64 | 2000 | 2.2464 |
|
176 |
+
| 2.1928 | 0.65 | 2025 | 2.2463 |
|
177 |
+
| 2.2087 | 0.66 | 2050 | 2.2455 |
|
178 |
+
| 2.1792 | 0.66 | 2075 | 2.2470 |
|
179 |
+
| 2.252 | 0.67 | 2100 | 2.2468 |
|
180 |
+
| 2.2018 | 0.68 | 2125 | 2.2456 |
|
181 |
+
| 2.2006 | 0.69 | 2150 | 2.2451 |
|
182 |
+
| 2.2076 | 0.7 | 2175 | 2.2449 |
|
183 |
+
| 2.2436 | 0.7 | 2200 | 2.2460 |
|
184 |
+
| 2.2156 | 0.71 | 2225 | 2.2477 |
|
185 |
+
| 2.1348 | 0.72 | 2250 | 2.2455 |
|
186 |
+
| 2.1338 | 0.73 | 2275 | 2.2450 |
|
187 |
+
| 2.2147 | 0.74 | 2300 | 2.2455 |
|
188 |
+
| 2.2766 | 0.74 | 2325 | 2.2444 |
|
189 |
+
| 2.204 | 0.75 | 2350 | 2.2458 |
|
190 |
+
|
191 |
+
|
192 |
+
### Framework versions
|
193 |
+
|
194 |
+
- Transformers 4.36.2
|
195 |
+
- Pytorch 2.1.2+cu121
|
196 |
+
- Datasets 2.15.0
|
197 |
+
- Tokenizers 0.15.0
|
198 |
+
|
199 |
+
|