End of training
Browse files- README.md +3 -2
- all_results.json +5 -0
- config.json +4 -3
- eval_results.json +5 -5
README.md
CHANGED
@@ -2,12 +2,13 @@
|
|
2 |
license: apache-2.0
|
3 |
library_name: peft
|
4 |
tags:
|
|
|
5 |
- trl
|
6 |
- sft
|
7 |
- generated_from_trainer
|
8 |
base_model: mistralai/Mistral-7B-v0.1
|
9 |
datasets:
|
10 |
-
-
|
11 |
model-index:
|
12 |
- name: zephyr-7b-sft-qlora
|
13 |
results: []
|
@@ -18,7 +19,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
18 |
|
19 |
# zephyr-7b-sft-qlora
|
20 |
|
21 |
-
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the
|
22 |
It achieves the following results on the evaluation set:
|
23 |
- Loss: 0.9517
|
24 |
|
|
|
2 |
license: apache-2.0
|
3 |
library_name: peft
|
4 |
tags:
|
5 |
+
- alignment-handbook
|
6 |
- trl
|
7 |
- sft
|
8 |
- generated_from_trainer
|
9 |
base_model: mistralai/Mistral-7B-v0.1
|
10 |
datasets:
|
11 |
+
- HuggingFaceH4/ultrachat_200k
|
12 |
model-index:
|
13 |
- name: zephyr-7b-sft-qlora
|
14 |
results: []
|
|
|
19 |
|
20 |
# zephyr-7b-sft-qlora
|
21 |
|
22 |
+
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the HuggingFaceH4/ultrachat_200k dataset.
|
23 |
It achieves the following results on the evaluation set:
|
24 |
- Loss: 0.9517
|
25 |
|
all_results.json
CHANGED
@@ -1,5 +1,10 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
|
|
|
|
|
|
|
|
|
|
3 |
"train_loss": 0.8657062510599203,
|
4 |
"train_runtime": 31020.7888,
|
5 |
"train_samples": 207864,
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"eval_loss": 0.9517214298248291,
|
4 |
+
"eval_runtime": 970.177,
|
5 |
+
"eval_samples": 23109,
|
6 |
+
"eval_samples_per_second": 15.905,
|
7 |
+
"eval_steps_per_second": 0.995,
|
8 |
"train_loss": 0.8657062510599203,
|
9 |
"train_runtime": 31020.7888,
|
10 |
"train_samples": 207864,
|
config.json
CHANGED
@@ -18,7 +18,8 @@
|
|
18 |
"quantization_config": {
|
19 |
"_load_in_4bit": true,
|
20 |
"_load_in_8bit": false,
|
21 |
-
"bnb_4bit_compute_dtype": "
|
|
|
22 |
"bnb_4bit_quant_type": "nf4",
|
23 |
"bnb_4bit_use_double_quant": false,
|
24 |
"llm_int8_enable_fp32_cpu_offload": false,
|
@@ -33,8 +34,8 @@
|
|
33 |
"rope_theta": 10000.0,
|
34 |
"sliding_window": 4096,
|
35 |
"tie_word_embeddings": false,
|
36 |
-
"torch_dtype": "
|
37 |
-
"transformers_version": "4.
|
38 |
"use_cache": true,
|
39 |
"vocab_size": 32000
|
40 |
}
|
|
|
18 |
"quantization_config": {
|
19 |
"_load_in_4bit": true,
|
20 |
"_load_in_8bit": false,
|
21 |
+
"bnb_4bit_compute_dtype": "bfloat16",
|
22 |
+
"bnb_4bit_quant_storage": "uint8",
|
23 |
"bnb_4bit_quant_type": "nf4",
|
24 |
"bnb_4bit_use_double_quant": false,
|
25 |
"llm_int8_enable_fp32_cpu_offload": false,
|
|
|
34 |
"rope_theta": 10000.0,
|
35 |
"sliding_window": 4096,
|
36 |
"tie_word_embeddings": false,
|
37 |
+
"torch_dtype": "bfloat16",
|
38 |
+
"transformers_version": "4.39.3",
|
39 |
"use_cache": true,
|
40 |
"vocab_size": 32000
|
41 |
}
|
eval_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"eval_loss": 0.
|
4 |
-
"eval_runtime":
|
5 |
-
"eval_samples":
|
6 |
-
"eval_samples_per_second": 15.
|
7 |
-
"eval_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"eval_loss": 0.9517214298248291,
|
4 |
+
"eval_runtime": 970.177,
|
5 |
+
"eval_samples": 23109,
|
6 |
+
"eval_samples_per_second": 15.905,
|
7 |
+
"eval_steps_per_second": 0.995
|
8 |
}
|