GaetanMichelet commited on
Commit
eda5629
1 Parent(s): 635b3b4

End of training

Browse files
README.md CHANGED
@@ -1,8 +1,12 @@
1
  ---
2
  base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
 
 
 
3
  library_name: peft
4
  license: llama3.1
5
  tags:
 
6
  - trl
7
  - sft
8
  - generated_from_trainer
@@ -16,9 +20,9 @@ should probably proofread and complete it, then remove this comment. -->
16
 
17
  # Llama-31-8B_task-1_120-samples_config-1_full
18
 
19
- This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct) on the None dataset.
20
  It achieves the following results on the evaluation set:
21
- - Loss: 1.5564
22
 
23
  ## Model description
24
 
 
1
  ---
2
  base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
3
+ datasets:
4
+ - GaetanMichelet/chat-60_ft_task-1
5
+ - GaetanMichelet/chat-120_ft_task-1
6
  library_name: peft
7
  license: llama3.1
8
  tags:
9
+ - alignment-handbook
10
  - trl
11
  - sft
12
  - generated_from_trainer
 
20
 
21
  # Llama-31-8B_task-1_120-samples_config-1_full
22
 
23
+ This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct) on the GaetanMichelet/chat-60_ft_task-1 and the GaetanMichelet/chat-120_ft_task-1 datasets.
24
  It achieves the following results on the evaluation set:
25
+ - Loss: 0.8656
26
 
27
  ## Model description
28
 
all_results.json CHANGED
@@ -1,5 +1,10 @@
1
  {
2
  "epoch": 14.0,
 
 
 
 
 
3
  "total_flos": 1.8910971254800384e+16,
4
  "train_loss": 0.8254564341593098,
5
  "train_runtime": 1444.0259,
 
1
  {
2
  "epoch": 14.0,
3
+ "eval_loss": 0.8656427264213562,
4
+ "eval_runtime": 9.6287,
5
+ "eval_samples": 24,
6
+ "eval_samples_per_second": 2.493,
7
+ "eval_steps_per_second": 2.493,
8
  "total_flos": 1.8910971254800384e+16,
9
  "train_loss": 0.8254564341593098,
10
  "train_runtime": 1444.0259,
config.json ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "meta-llama/Meta-Llama-3.1-8B-Instruct",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 128000,
9
+ "eos_token_id": [
10
+ 128001,
11
+ 128008,
12
+ 128009
13
+ ],
14
+ "hidden_act": "silu",
15
+ "hidden_size": 4096,
16
+ "initializer_range": 0.02,
17
+ "intermediate_size": 14336,
18
+ "max_position_embeddings": 131072,
19
+ "mlp_bias": false,
20
+ "model_type": "llama",
21
+ "num_attention_heads": 32,
22
+ "num_hidden_layers": 32,
23
+ "num_key_value_heads": 8,
24
+ "pretraining_tp": 1,
25
+ "quantization_config": {
26
+ "_load_in_4bit": true,
27
+ "_load_in_8bit": false,
28
+ "bnb_4bit_compute_dtype": "bfloat16",
29
+ "bnb_4bit_quant_storage": "uint8",
30
+ "bnb_4bit_quant_type": "nf4",
31
+ "bnb_4bit_use_double_quant": false,
32
+ "llm_int8_enable_fp32_cpu_offload": false,
33
+ "llm_int8_has_fp16_weight": false,
34
+ "llm_int8_skip_modules": null,
35
+ "llm_int8_threshold": 6.0,
36
+ "load_in_4bit": true,
37
+ "load_in_8bit": false,
38
+ "quant_method": "bitsandbytes"
39
+ },
40
+ "rms_norm_eps": 1e-05,
41
+ "rope_scaling": {
42
+ "factor": 8.0,
43
+ "high_freq_factor": 4.0,
44
+ "low_freq_factor": 1.0,
45
+ "original_max_position_embeddings": 8192,
46
+ "rope_type": "llama3"
47
+ },
48
+ "rope_theta": 500000.0,
49
+ "tie_word_embeddings": false,
50
+ "torch_dtype": "bfloat16",
51
+ "transformers_version": "4.44.0",
52
+ "use_cache": true,
53
+ "vocab_size": 128256
54
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 14.0,
3
+ "eval_loss": 0.8656427264213562,
4
+ "eval_runtime": 9.6287,
5
+ "eval_samples": 24,
6
+ "eval_samples_per_second": 2.493,
7
+ "eval_steps_per_second": 2.493
8
+ }
runs/Aug20_14-49-24_9mqd6r3-l/events.out.tfevents.1724159666.9mqd6r3-l.1482162.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a03a99a0ae192b2c6add2a545fd4158d08db497a82af89cd5cd7674d5ae82339
3
+ size 359