GaetanMichelet
commited on
Commit
•
6aa190b
1
Parent(s):
93115b5
End of training
Browse files- README.md +5 -2
- all_results.json +5 -0
- config.json +42 -0
- eval_results.json +8 -0
- runs/Aug24_07-12-33_9mqd6r3-l/events.out.tfevents.1724479089.9mqd6r3-l.2888508.1 +3 -0
README.md
CHANGED
@@ -1,8 +1,11 @@
|
|
1 |
---
|
2 |
base_model: mistralai/Mistral-7B-Instruct-v0.3
|
|
|
|
|
3 |
library_name: peft
|
4 |
license: apache-2.0
|
5 |
tags:
|
|
|
6 |
- trl
|
7 |
- sft
|
8 |
- generated_from_trainer
|
@@ -16,9 +19,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
16 |
|
17 |
# Mistral-7B_task-2_60-samples_config-2_auto
|
18 |
|
19 |
-
This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.3](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3) on
|
20 |
It achieves the following results on the evaluation set:
|
21 |
-
- Loss: 0.
|
22 |
|
23 |
## Model description
|
24 |
|
|
|
1 |
---
|
2 |
base_model: mistralai/Mistral-7B-Instruct-v0.3
|
3 |
+
datasets:
|
4 |
+
- GaetanMichelet/chat-60_ft_task-2_auto
|
5 |
library_name: peft
|
6 |
license: apache-2.0
|
7 |
tags:
|
8 |
+
- alignment-handbook
|
9 |
- trl
|
10 |
- sft
|
11 |
- generated_from_trainer
|
|
|
19 |
|
20 |
# Mistral-7B_task-2_60-samples_config-2_auto
|
21 |
|
22 |
+
This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.3](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3) on the GaetanMichelet/chat-60_ft_task-2_auto dataset.
|
23 |
It achieves the following results on the evaluation set:
|
24 |
+
- Loss: 0.4866
|
25 |
|
26 |
## Model description
|
27 |
|
all_results.json
CHANGED
@@ -1,5 +1,10 @@
|
|
1 |
{
|
2 |
"epoch": 12.869565217391305,
|
|
|
|
|
|
|
|
|
|
|
3 |
"total_flos": 7.417401732130406e+16,
|
4 |
"train_loss": 0.3416548876746281,
|
5 |
"train_runtime": 2693.8062,
|
|
|
1 |
{
|
2 |
"epoch": 12.869565217391305,
|
3 |
+
"eval_loss": 0.48655232787132263,
|
4 |
+
"eval_runtime": 16.5159,
|
5 |
+
"eval_samples": 12,
|
6 |
+
"eval_samples_per_second": 0.727,
|
7 |
+
"eval_steps_per_second": 0.727,
|
8 |
"total_flos": 7.417401732130406e+16,
|
9 |
"train_loss": 0.3416548876746281,
|
10 |
"train_runtime": 2693.8062,
|
config.json
ADDED
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "mistralai/Mistral-7B-Instruct-v0.3",
|
3 |
+
"architectures": [
|
4 |
+
"MistralForCausalLM"
|
5 |
+
],
|
6 |
+
"attention_dropout": 0.0,
|
7 |
+
"bos_token_id": 1,
|
8 |
+
"eos_token_id": 2,
|
9 |
+
"head_dim": 128,
|
10 |
+
"hidden_act": "silu",
|
11 |
+
"hidden_size": 4096,
|
12 |
+
"initializer_range": 0.02,
|
13 |
+
"intermediate_size": 14336,
|
14 |
+
"max_position_embeddings": 32768,
|
15 |
+
"model_type": "mistral",
|
16 |
+
"num_attention_heads": 32,
|
17 |
+
"num_hidden_layers": 32,
|
18 |
+
"num_key_value_heads": 8,
|
19 |
+
"quantization_config": {
|
20 |
+
"_load_in_4bit": true,
|
21 |
+
"_load_in_8bit": false,
|
22 |
+
"bnb_4bit_compute_dtype": "bfloat16",
|
23 |
+
"bnb_4bit_quant_storage": "uint8",
|
24 |
+
"bnb_4bit_quant_type": "nf4",
|
25 |
+
"bnb_4bit_use_double_quant": false,
|
26 |
+
"llm_int8_enable_fp32_cpu_offload": false,
|
27 |
+
"llm_int8_has_fp16_weight": false,
|
28 |
+
"llm_int8_skip_modules": null,
|
29 |
+
"llm_int8_threshold": 6.0,
|
30 |
+
"load_in_4bit": true,
|
31 |
+
"load_in_8bit": false,
|
32 |
+
"quant_method": "bitsandbytes"
|
33 |
+
},
|
34 |
+
"rms_norm_eps": 1e-05,
|
35 |
+
"rope_theta": 1000000.0,
|
36 |
+
"sliding_window": null,
|
37 |
+
"tie_word_embeddings": false,
|
38 |
+
"torch_dtype": "bfloat16",
|
39 |
+
"transformers_version": "4.44.0",
|
40 |
+
"use_cache": true,
|
41 |
+
"vocab_size": 32768
|
42 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 12.869565217391305,
|
3 |
+
"eval_loss": 0.48655232787132263,
|
4 |
+
"eval_runtime": 16.5159,
|
5 |
+
"eval_samples": 12,
|
6 |
+
"eval_samples_per_second": 0.727,
|
7 |
+
"eval_steps_per_second": 0.727
|
8 |
+
}
|
runs/Aug24_07-12-33_9mqd6r3-l/events.out.tfevents.1724479089.9mqd6r3-l.2888508.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fa7237cf896d4fcab0e5791fe120aa5108c48839a167960a903f0ecd5c4917c2
|
3 |
+
size 354
|