Batch upload part 15
Browse files- nl_tasks/exp395/run_ex03/ft/adapter_config.json +18 -0
- nl_tasks/exp395/run_ex03/ft/tokenizer.json +0 -0
- nl_tasks/exp395/run_ex03/ft/tokenizer.model +3 -0
- nl_tasks/exp395/run_ex03/ft2/adapter_config.json +18 -0
- nl_tasks/exp395/run_ex03/ft2/adapter_model.bin +3 -0
- nl_tasks/exp395/run_ex04/ft/adapter_config.json +18 -0
- nl_tasks/exp395/run_ex04/ft/added_tokens.json +3 -0
- nl_tasks/exp395/run_ex04/ft/special_tokens_map.json +30 -0
- nl_tasks/exp395/run_ex04/ft/tokenizer.json +0 -0
- nl_tasks/exp395/run_ex04/ft/tokenizer.model +3 -0
- nl_tasks/exp395/run_ex04/ft/tokenizer_config.json +51 -0
- nl_tasks/exp395/run_ex04/ft2/adapter_config.json +18 -0
- nl_tasks/exp395/run_ex04/ft2/adapter_model.bin +3 -0
- nl_tasks/exp395/run_ex04/trainer_state.json +308 -0
- nl_tasks/exp395/run_ex05/ft/adapter_config.json +18 -0
- nl_tasks/exp395/run_ex05/ft/added_tokens.json +3 -0
- nl_tasks/exp395/run_ex05/ft/special_tokens_map.json +30 -0
- nl_tasks/exp395/run_ex05/ft/tokenizer.json +0 -0
- nl_tasks/exp395/run_ex05/ft/tokenizer.model +3 -0
- nl_tasks/exp395/run_ex05/ft/tokenizer_config.json +51 -0
- nl_tasks/exp395/run_ex05/ft2/adapter_config.json +18 -0
- nl_tasks/exp395/run_ex05/ft2/adapter_model.bin +3 -0
- nl_tasks/exp395/run_ex05/trainer_state.json +509 -0
- nl_tasks/exp395/run_ex06/ft/adapter_config.json +18 -0
- nl_tasks/exp395/run_ex06/ft/added_tokens.json +3 -0
- nl_tasks/exp395/run_ex06/ft/special_tokens_map.json +30 -0
- nl_tasks/exp395/run_ex06/ft/tokenizer.json +0 -0
- nl_tasks/exp395/run_ex06/ft/tokenizer.model +3 -0
- nl_tasks/exp395/run_ex06/ft/tokenizer_config.json +51 -0
- nl_tasks/exp395/run_ex06/ft2/adapter_config.json +18 -0
- nl_tasks/exp395/run_ex06/ft2/adapter_model.bin +3 -0
- nl_tasks/exp395/run_ex06/trainer_state.json +509 -0
- nl_tasks/exp395/run_ex07/ft/adapter_config.json +18 -0
- nl_tasks/exp395/run_ex07/ft/added_tokens.json +3 -0
- nl_tasks/exp395/run_ex07/ft/special_tokens_map.json +30 -0
- nl_tasks/exp395/run_ex07/ft/tokenizer.json +0 -0
- nl_tasks/exp395/run_ex07/ft/tokenizer.model +3 -0
- nl_tasks/exp395/run_ex07/ft/tokenizer_config.json +51 -0
- nl_tasks/exp395/run_ex07/ft2/adapter_config.json +18 -0
- nl_tasks/exp395/run_ex07/ft2/adapter_model.bin +3 -0
- nl_tasks/exp395/run_ex07/trainer_state.json +509 -0
- nl_tasks/exp395/run_ex08/ft/adapter_config.json +18 -0
- nl_tasks/exp395/run_ex08/ft/added_tokens.json +3 -0
- nl_tasks/exp395/run_ex08/ft/special_tokens_map.json +30 -0
- nl_tasks/exp395/run_ex08/ft/tokenizer.json +0 -0
- nl_tasks/exp395/run_ex08/ft/tokenizer.model +3 -0
- nl_tasks/exp395/run_ex08/ft/tokenizer_config.json +51 -0
- nl_tasks/exp395/run_ex08/ft2/adapter_config.json +18 -0
- nl_tasks/exp395/run_ex08/ft2/adapter_model.bin +3 -0
- nl_tasks/exp395/run_ex08/trainer_state.json +509 -0
nl_tasks/exp395/run_ex03/ft/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"T": 1.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"inference_mode": false,
|
| 6 |
+
"layers_to_transform": null,
|
| 7 |
+
"modules_to_save": null,
|
| 8 |
+
"num_rotations": 1,
|
| 9 |
+
"peft_type": "ROTATION",
|
| 10 |
+
"r": 16,
|
| 11 |
+
"revision": null,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"q_proj",
|
| 14 |
+
"v_proj"
|
| 15 |
+
],
|
| 16 |
+
"target_modules_to_skip": null,
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
nl_tasks/exp395/run_ex03/ft/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
nl_tasks/exp395/run_ex03/ft/tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
nl_tasks/exp395/run_ex03/ft2/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"T": 1.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"layers_to_transform": null,
|
| 7 |
+
"modules_to_save": null,
|
| 8 |
+
"num_rotations": 1,
|
| 9 |
+
"peft_type": "ROTATION",
|
| 10 |
+
"r": 16,
|
| 11 |
+
"revision": null,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"q_proj",
|
| 14 |
+
"v_proj"
|
| 15 |
+
],
|
| 16 |
+
"target_modules_to_skip": null,
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
nl_tasks/exp395/run_ex03/ft2/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:503674d6e9d395776c6c79b8d13ca3ea16bd92520277eed1c79858595d63572f
|
| 3 |
+
size 33602915
|
nl_tasks/exp395/run_ex04/ft/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"T": 1.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"inference_mode": false,
|
| 6 |
+
"layers_to_transform": null,
|
| 7 |
+
"modules_to_save": null,
|
| 8 |
+
"num_rotations": 1,
|
| 9 |
+
"peft_type": "ROTATION",
|
| 10 |
+
"r": 16,
|
| 11 |
+
"revision": null,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"v_proj",
|
| 14 |
+
"q_proj"
|
| 15 |
+
],
|
| 16 |
+
"target_modules_to_skip": null,
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
nl_tasks/exp395/run_ex04/ft/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[PAD]": 32000
|
| 3 |
+
}
|
nl_tasks/exp395/run_ex04/ft/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"pad_token": {
|
| 17 |
+
"content": "[PAD]",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "</s>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
nl_tasks/exp395/run_ex04/ft/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
nl_tasks/exp395/run_ex04/ft/tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
nl_tasks/exp395/run_ex04/ft/tokenizer_config.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": null,
|
| 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 |
+
"32000": {
|
| 31 |
+
"content": "[PAD]",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false,
|
| 36 |
+
"special": true
|
| 37 |
+
}
|
| 38 |
+
},
|
| 39 |
+
"bos_token": "</s>",
|
| 40 |
+
"clean_up_tokenization_spaces": false,
|
| 41 |
+
"eos_token": "</s>",
|
| 42 |
+
"extra_special_tokens": {},
|
| 43 |
+
"legacy": false,
|
| 44 |
+
"model_max_length": 512,
|
| 45 |
+
"pad_token": "[PAD]",
|
| 46 |
+
"padding_side": "right",
|
| 47 |
+
"sp_model_kwargs": {},
|
| 48 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 49 |
+
"unk_token": "</s>",
|
| 50 |
+
"use_default_system_prompt": false
|
| 51 |
+
}
|
nl_tasks/exp395/run_ex04/ft2/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"T": 1.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"layers_to_transform": null,
|
| 7 |
+
"modules_to_save": null,
|
| 8 |
+
"num_rotations": 1,
|
| 9 |
+
"peft_type": "ROTATION",
|
| 10 |
+
"r": 16,
|
| 11 |
+
"revision": null,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"v_proj",
|
| 14 |
+
"q_proj"
|
| 15 |
+
],
|
| 16 |
+
"target_modules_to_skip": null,
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
nl_tasks/exp395/run_ex04/ft2/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37d6b88a9707e400bbde1c970e5d72b090db0d40ad030b4323c4c3941c0192e9
|
| 3 |
+
size 33602915
|
nl_tasks/exp395/run_ex04/trainer_state.json
ADDED
|
@@ -0,0 +1,308 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 1000,
|
| 7 |
+
"global_step": 6234,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.06416426050689766,
|
| 14 |
+
"grad_norm": 0.21872752904891968,
|
| 15 |
+
"learning_rate": 0.0009988020657471077,
|
| 16 |
+
"loss": 0.3629,
|
| 17 |
+
"step": 200
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.12832852101379533,
|
| 21 |
+
"grad_norm": 0.21329320967197418,
|
| 22 |
+
"learning_rate": 0.0009927029382680798,
|
| 23 |
+
"loss": 0.2964,
|
| 24 |
+
"step": 400
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.19249278152069296,
|
| 28 |
+
"grad_norm": 1.0523426532745361,
|
| 29 |
+
"learning_rate": 0.0009815004229628667,
|
| 30 |
+
"loss": 0.2791,
|
| 31 |
+
"step": 600
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.25665704202759065,
|
| 35 |
+
"grad_norm": 0.2172887623310089,
|
| 36 |
+
"learning_rate": 0.0009653105548208928,
|
| 37 |
+
"loss": 0.2617,
|
| 38 |
+
"step": 800
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.3208213025344883,
|
| 42 |
+
"grad_norm": 0.18885284662246704,
|
| 43 |
+
"learning_rate": 0.0009443010275366055,
|
| 44 |
+
"loss": 0.2506,
|
| 45 |
+
"step": 1000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.3208213025344883,
|
| 49 |
+
"eval_loss": 0.2593642771244049,
|
| 50 |
+
"eval_runtime": 5.2735,
|
| 51 |
+
"eval_samples_per_second": 48.924,
|
| 52 |
+
"eval_steps_per_second": 0.948,
|
| 53 |
+
"step": 1000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.3849855630413859,
|
| 57 |
+
"grad_norm": 0.15659989416599274,
|
| 58 |
+
"learning_rate": 0.0009186894565481294,
|
| 59 |
+
"loss": 0.2423,
|
| 60 |
+
"step": 1200
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.4491498235482836,
|
| 64 |
+
"grad_norm": 0.14614759385585785,
|
| 65 |
+
"learning_rate": 0.0008887411249894593,
|
| 66 |
+
"loss": 0.2359,
|
| 67 |
+
"step": 1400
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 0.5133140840551813,
|
| 71 |
+
"grad_norm": 0.13668683171272278,
|
| 72 |
+
"learning_rate": 0.0008547662359034838,
|
| 73 |
+
"loss": 0.2374,
|
| 74 |
+
"step": 1600
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.5774783445620789,
|
| 78 |
+
"grad_norm": 0.14054378867149353,
|
| 79 |
+
"learning_rate": 0.0008171166991772578,
|
| 80 |
+
"loss": 0.2273,
|
| 81 |
+
"step": 1800
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.6416426050689766,
|
| 85 |
+
"grad_norm": 0.11869219690561295,
|
| 86 |
+
"learning_rate": 0.0007761824864802529,
|
| 87 |
+
"loss": 0.2222,
|
| 88 |
+
"step": 2000
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 0.6416426050689766,
|
| 92 |
+
"eval_loss": 0.23084010183811188,
|
| 93 |
+
"eval_runtime": 5.0454,
|
| 94 |
+
"eval_samples_per_second": 51.136,
|
| 95 |
+
"eval_steps_per_second": 0.991,
|
| 96 |
+
"step": 2000
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.7058068655758742,
|
| 100 |
+
"grad_norm": 0.1382647603750229,
|
| 101 |
+
"learning_rate": 0.0007323875919609245,
|
| 102 |
+
"loss": 0.224,
|
| 103 |
+
"step": 2200
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.7699711260827719,
|
| 107 |
+
"grad_norm": 0.1192302256822586,
|
| 108 |
+
"learning_rate": 0.0006861856405404567,
|
| 109 |
+
"loss": 0.2161,
|
| 110 |
+
"step": 2400
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.8341353865896696,
|
| 114 |
+
"grad_norm": 0.12821270525455475,
|
| 115 |
+
"learning_rate": 0.0006380551892927205,
|
| 116 |
+
"loss": 0.2164,
|
| 117 |
+
"step": 2600
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.8982996470965672,
|
| 121 |
+
"grad_norm": 0.1226765364408493,
|
| 122 |
+
"learning_rate": 0.0005884947705784723,
|
| 123 |
+
"loss": 0.2119,
|
| 124 |
+
"step": 2800
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.9624639076034649,
|
| 128 |
+
"grad_norm": 0.11759891360998154,
|
| 129 |
+
"learning_rate": 0.0005380177282767159,
|
| 130 |
+
"loss": 0.2127,
|
| 131 |
+
"step": 3000
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.9624639076034649,
|
| 135 |
+
"eval_loss": 0.21615009009838104,
|
| 136 |
+
"eval_runtime": 5.0486,
|
| 137 |
+
"eval_samples_per_second": 51.103,
|
| 138 |
+
"eval_steps_per_second": 0.99,
|
| 139 |
+
"step": 3000
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 1.0266281681103626,
|
| 143 |
+
"grad_norm": 0.11819776147603989,
|
| 144 |
+
"learning_rate": 0.0004871469005992345,
|
| 145 |
+
"loss": 0.1981,
|
| 146 |
+
"step": 3200
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.0907924286172601,
|
| 150 |
+
"grad_norm": 0.11975960433483124,
|
| 151 |
+
"learning_rate": 0.00043640920456338925,
|
| 152 |
+
"loss": 0.1804,
|
| 153 |
+
"step": 3400
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 1.1549566891241578,
|
| 157 |
+
"grad_norm": 0.160000741481781,
|
| 158 |
+
"learning_rate": 0.00038633017821688954,
|
| 159 |
+
"loss": 0.1847,
|
| 160 |
+
"step": 3600
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 1.2191209496310556,
|
| 164 |
+
"grad_norm": 0.11288543790578842,
|
| 165 |
+
"learning_rate": 0.00033742853714584963,
|
| 166 |
+
"loss": 0.1821,
|
| 167 |
+
"step": 3800
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 1.283285210137953,
|
| 171 |
+
"grad_norm": 0.11609069257974625,
|
| 172 |
+
"learning_rate": 0.0002902108016494981,
|
| 173 |
+
"loss": 0.1803,
|
| 174 |
+
"step": 4000
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 1.283285210137953,
|
| 178 |
+
"eval_loss": 0.20613692700862885,
|
| 179 |
+
"eval_runtime": 5.0502,
|
| 180 |
+
"eval_samples_per_second": 51.087,
|
| 181 |
+
"eval_steps_per_second": 0.99,
|
| 182 |
+
"step": 4000
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.3474494706448508,
|
| 186 |
+
"grad_norm": 0.12499521672725677,
|
| 187 |
+
"learning_rate": 0.0002451660502329462,
|
| 188 |
+
"loss": 0.1827,
|
| 189 |
+
"step": 4200
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.4116137311517485,
|
| 193 |
+
"grad_norm": 0.13932673633098602,
|
| 194 |
+
"learning_rate": 0.0002027608537610257,
|
| 195 |
+
"loss": 0.1758,
|
| 196 |
+
"step": 4400
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.4757779916586462,
|
| 200 |
+
"grad_norm": 0.10319329053163528,
|
| 201 |
+
"learning_rate": 0.00016343444274492946,
|
| 202 |
+
"loss": 0.1796,
|
| 203 |
+
"step": 4600
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.539942252165544,
|
| 207 |
+
"grad_norm": 0.09759525954723358,
|
| 208 |
+
"learning_rate": 0.00012759415781861554,
|
| 209 |
+
"loss": 0.176,
|
| 210 |
+
"step": 4800
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.6041065126724414,
|
| 214 |
+
"grad_norm": 0.10206837952136993,
|
| 215 |
+
"learning_rate": 9.561123052866844e-05,
|
| 216 |
+
"loss": 0.1755,
|
| 217 |
+
"step": 5000
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.6041065126724414,
|
| 221 |
+
"eval_loss": 0.19889786839485168,
|
| 222 |
+
"eval_runtime": 5.0526,
|
| 223 |
+
"eval_samples_per_second": 51.063,
|
| 224 |
+
"eval_steps_per_second": 0.99,
|
| 225 |
+
"step": 5000
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 1.6682707731793391,
|
| 229 |
+
"grad_norm": 0.0990043506026268,
|
| 230 |
+
"learning_rate": 6.781693813994666e-05,
|
| 231 |
+
"loss": 0.1747,
|
| 232 |
+
"step": 5200
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 1.7324350336862366,
|
| 236 |
+
"grad_norm": 0.11463545262813568,
|
| 237 |
+
"learning_rate": 4.449917228531264e-05,
|
| 238 |
+
"loss": 0.1741,
|
| 239 |
+
"step": 5400
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.7965992941931344,
|
| 243 |
+
"grad_norm": 0.1105906218290329,
|
| 244 |
+
"learning_rate": 2.5899457001165806e-05,
|
| 245 |
+
"loss": 0.1745,
|
| 246 |
+
"step": 5600
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 1.860763554700032,
|
| 250 |
+
"grad_norm": 0.09413338452577591,
|
| 251 |
+
"learning_rate": 1.2210447035791717e-05,
|
| 252 |
+
"loss": 0.174,
|
| 253 |
+
"step": 5800
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 1.9249278152069298,
|
| 257 |
+
"grad_norm": 0.11482080072164536,
|
| 258 |
+
"learning_rate": 3.5739323429003765e-06,
|
| 259 |
+
"loss": 0.1699,
|
| 260 |
+
"step": 6000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.9249278152069298,
|
| 264 |
+
"eval_loss": 0.19647078216075897,
|
| 265 |
+
"eval_runtime": 5.0578,
|
| 266 |
+
"eval_samples_per_second": 51.011,
|
| 267 |
+
"eval_steps_per_second": 0.989,
|
| 268 |
+
"step": 6000
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.9890920757138275,
|
| 272 |
+
"grad_norm": 0.11400256305932999,
|
| 273 |
+
"learning_rate": 7.936942969133387e-08,
|
| 274 |
+
"loss": 0.172,
|
| 275 |
+
"step": 6200
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 2.0,
|
| 279 |
+
"step": 6234,
|
| 280 |
+
"total_flos": 4.0542214473488794e+18,
|
| 281 |
+
"train_loss": 0.21111015283721052,
|
| 282 |
+
"train_runtime": 5747.5761,
|
| 283 |
+
"train_samples_per_second": 34.708,
|
| 284 |
+
"train_steps_per_second": 1.085
|
| 285 |
+
}
|
| 286 |
+
],
|
| 287 |
+
"logging_steps": 200,
|
| 288 |
+
"max_steps": 6234,
|
| 289 |
+
"num_input_tokens_seen": 0,
|
| 290 |
+
"num_train_epochs": 2,
|
| 291 |
+
"save_steps": 0,
|
| 292 |
+
"stateful_callbacks": {
|
| 293 |
+
"TrainerControl": {
|
| 294 |
+
"args": {
|
| 295 |
+
"should_epoch_stop": false,
|
| 296 |
+
"should_evaluate": false,
|
| 297 |
+
"should_log": false,
|
| 298 |
+
"should_save": true,
|
| 299 |
+
"should_training_stop": true
|
| 300 |
+
},
|
| 301 |
+
"attributes": {}
|
| 302 |
+
}
|
| 303 |
+
},
|
| 304 |
+
"total_flos": 4.0542214473488794e+18,
|
| 305 |
+
"train_batch_size": 32,
|
| 306 |
+
"trial_name": null,
|
| 307 |
+
"trial_params": null
|
| 308 |
+
}
|
nl_tasks/exp395/run_ex05/ft/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"T": 1.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"inference_mode": false,
|
| 6 |
+
"layers_to_transform": null,
|
| 7 |
+
"modules_to_save": null,
|
| 8 |
+
"num_rotations": 1,
|
| 9 |
+
"peft_type": "ROTATION",
|
| 10 |
+
"r": 16,
|
| 11 |
+
"revision": null,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"v_proj",
|
| 14 |
+
"q_proj"
|
| 15 |
+
],
|
| 16 |
+
"target_modules_to_skip": null,
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
nl_tasks/exp395/run_ex05/ft/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[PAD]": 32000
|
| 3 |
+
}
|
nl_tasks/exp395/run_ex05/ft/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"pad_token": {
|
| 17 |
+
"content": "[PAD]",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "</s>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
nl_tasks/exp395/run_ex05/ft/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
nl_tasks/exp395/run_ex05/ft/tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
nl_tasks/exp395/run_ex05/ft/tokenizer_config.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": null,
|
| 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 |
+
"32000": {
|
| 31 |
+
"content": "[PAD]",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false,
|
| 36 |
+
"special": true
|
| 37 |
+
}
|
| 38 |
+
},
|
| 39 |
+
"bos_token": "</s>",
|
| 40 |
+
"clean_up_tokenization_spaces": false,
|
| 41 |
+
"eos_token": "</s>",
|
| 42 |
+
"extra_special_tokens": {},
|
| 43 |
+
"legacy": false,
|
| 44 |
+
"model_max_length": 512,
|
| 45 |
+
"pad_token": "[PAD]",
|
| 46 |
+
"padding_side": "right",
|
| 47 |
+
"sp_model_kwargs": {},
|
| 48 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 49 |
+
"unk_token": "</s>",
|
| 50 |
+
"use_default_system_prompt": false
|
| 51 |
+
}
|
nl_tasks/exp395/run_ex05/ft2/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"T": 1.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"layers_to_transform": null,
|
| 7 |
+
"modules_to_save": null,
|
| 8 |
+
"num_rotations": 1,
|
| 9 |
+
"peft_type": "ROTATION",
|
| 10 |
+
"r": 16,
|
| 11 |
+
"revision": null,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"v_proj",
|
| 14 |
+
"q_proj"
|
| 15 |
+
],
|
| 16 |
+
"target_modules_to_skip": null,
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
nl_tasks/exp395/run_ex05/ft2/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:946d5729b3750ab24b3fb2901f0a03fbdd1387b84bb653831d0180eabbdd6639
|
| 3 |
+
size 33602915
|
nl_tasks/exp395/run_ex05/trainer_state.json
ADDED
|
@@ -0,0 +1,509 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 9375,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.064,
|
| 14 |
+
"grad_norm": 0.19703811407089233,
|
| 15 |
+
"learning_rate": 0.000999684221114305,
|
| 16 |
+
"loss": 0.3657,
|
| 17 |
+
"step": 200
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.128,
|
| 21 |
+
"grad_norm": 0.2437756508588791,
|
| 22 |
+
"learning_rate": 0.0009973376564462874,
|
| 23 |
+
"loss": 0.2968,
|
| 24 |
+
"step": 400
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.16,
|
| 28 |
+
"eval_loss": 0.28636598587036133,
|
| 29 |
+
"eval_runtime": 19.6604,
|
| 30 |
+
"eval_samples_per_second": 51.423,
|
| 31 |
+
"eval_steps_per_second": 0.814,
|
| 32 |
+
"step": 500
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.192,
|
| 36 |
+
"grad_norm": 0.19999830424785614,
|
| 37 |
+
"learning_rate": 0.0009927125570277145,
|
| 38 |
+
"loss": 0.2773,
|
| 39 |
+
"step": 600
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.256,
|
| 43 |
+
"grad_norm": 0.20367808640003204,
|
| 44 |
+
"learning_rate": 0.0009858301125867587,
|
| 45 |
+
"loss": 0.2633,
|
| 46 |
+
"step": 800
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.32,
|
| 50 |
+
"grad_norm": 0.18598036468029022,
|
| 51 |
+
"learning_rate": 0.000976721854797373,
|
| 52 |
+
"loss": 0.2516,
|
| 53 |
+
"step": 1000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.32,
|
| 57 |
+
"eval_loss": 0.254931777715683,
|
| 58 |
+
"eval_runtime": 19.5103,
|
| 59 |
+
"eval_samples_per_second": 51.819,
|
| 60 |
+
"eval_steps_per_second": 0.82,
|
| 61 |
+
"step": 1000
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.384,
|
| 65 |
+
"grad_norm": 0.1677393764257431,
|
| 66 |
+
"learning_rate": 0.0009654295128180494,
|
| 67 |
+
"loss": 0.2477,
|
| 68 |
+
"step": 1200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.448,
|
| 72 |
+
"grad_norm": 0.15132741630077362,
|
| 73 |
+
"learning_rate": 0.0009520048221111679,
|
| 74 |
+
"loss": 0.2402,
|
| 75 |
+
"step": 1400
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.48,
|
| 79 |
+
"eval_loss": 0.2429439127445221,
|
| 80 |
+
"eval_runtime": 19.4737,
|
| 81 |
+
"eval_samples_per_second": 51.916,
|
| 82 |
+
"eval_steps_per_second": 0.822,
|
| 83 |
+
"step": 1500
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.512,
|
| 87 |
+
"grad_norm": 0.15494993329048157,
|
| 88 |
+
"learning_rate": 0.0009365092874188177,
|
| 89 |
+
"loss": 0.2342,
|
| 90 |
+
"step": 1600
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.576,
|
| 94 |
+
"grad_norm": 0.13975529372692108,
|
| 95 |
+
"learning_rate": 0.0009190139009810141,
|
| 96 |
+
"loss": 0.2298,
|
| 97 |
+
"step": 1800
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.64,
|
| 101 |
+
"grad_norm": 0.12113900482654572,
|
| 102 |
+
"learning_rate": 0.0008995988172872798,
|
| 103 |
+
"loss": 0.2279,
|
| 104 |
+
"step": 2000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.64,
|
| 108 |
+
"eval_loss": 0.2299424409866333,
|
| 109 |
+
"eval_runtime": 19.5012,
|
| 110 |
+
"eval_samples_per_second": 51.843,
|
| 111 |
+
"eval_steps_per_second": 0.82,
|
| 112 |
+
"step": 2000
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.704,
|
| 116 |
+
"grad_norm": 0.10823472589254379,
|
| 117 |
+
"learning_rate": 0.0008783529858517077,
|
| 118 |
+
"loss": 0.2234,
|
| 119 |
+
"step": 2200
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.768,
|
| 123 |
+
"grad_norm": 0.10321851819753647,
|
| 124 |
+
"learning_rate": 0.0008553737436939324,
|
| 125 |
+
"loss": 0.2215,
|
| 126 |
+
"step": 2400
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.8,
|
| 130 |
+
"eval_loss": 0.2244570553302765,
|
| 131 |
+
"eval_runtime": 19.4699,
|
| 132 |
+
"eval_samples_per_second": 51.926,
|
| 133 |
+
"eval_steps_per_second": 0.822,
|
| 134 |
+
"step": 2500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.832,
|
| 138 |
+
"grad_norm": 0.12440883368253708,
|
| 139 |
+
"learning_rate": 0.0008307663693930426,
|
| 140 |
+
"loss": 0.219,
|
| 141 |
+
"step": 2600
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.896,
|
| 145 |
+
"grad_norm": 0.10718903690576553,
|
| 146 |
+
"learning_rate": 0.000804643600757522,
|
| 147 |
+
"loss": 0.2182,
|
| 148 |
+
"step": 2800
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.96,
|
| 152 |
+
"grad_norm": 0.10340970754623413,
|
| 153 |
+
"learning_rate": 0.0007771251183209993,
|
| 154 |
+
"loss": 0.2114,
|
| 155 |
+
"step": 3000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.96,
|
| 159 |
+
"eval_loss": 0.220004603266716,
|
| 160 |
+
"eval_runtime": 19.5042,
|
| 161 |
+
"eval_samples_per_second": 51.835,
|
| 162 |
+
"eval_steps_per_second": 0.82,
|
| 163 |
+
"step": 3000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.024,
|
| 167 |
+
"grad_norm": 0.17328748106956482,
|
| 168 |
+
"learning_rate": 0.0007483369970301455,
|
| 169 |
+
"loss": 0.2057,
|
| 170 |
+
"step": 3200
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.088,
|
| 174 |
+
"grad_norm": 0.09701967239379883,
|
| 175 |
+
"learning_rate": 0.0007184111286368001,
|
| 176 |
+
"loss": 0.189,
|
| 177 |
+
"step": 3400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.12,
|
| 181 |
+
"eval_loss": 0.21541637182235718,
|
| 182 |
+
"eval_runtime": 19.5442,
|
| 183 |
+
"eval_samples_per_second": 51.729,
|
| 184 |
+
"eval_steps_per_second": 0.819,
|
| 185 |
+
"step": 3500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.152,
|
| 189 |
+
"grad_norm": 0.10150015354156494,
|
| 190 |
+
"learning_rate": 0.0006874846174406093,
|
| 191 |
+
"loss": 0.1891,
|
| 192 |
+
"step": 3600
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.216,
|
| 196 |
+
"grad_norm": 0.10405285656452179,
|
| 197 |
+
"learning_rate": 0.0006556991521505633,
|
| 198 |
+
"loss": 0.1891,
|
| 199 |
+
"step": 3800
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.28,
|
| 203 |
+
"grad_norm": 0.09167412668466568,
|
| 204 |
+
"learning_rate": 0.0006232003567432241,
|
| 205 |
+
"loss": 0.1869,
|
| 206 |
+
"step": 4000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.28,
|
| 210 |
+
"eval_loss": 0.21192973852157593,
|
| 211 |
+
"eval_runtime": 19.5118,
|
| 212 |
+
"eval_samples_per_second": 51.815,
|
| 213 |
+
"eval_steps_per_second": 0.82,
|
| 214 |
+
"step": 4000
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.3439999999999999,
|
| 218 |
+
"grad_norm": 0.09688983112573624,
|
| 219 |
+
"learning_rate": 0.0005901371232916676,
|
| 220 |
+
"loss": 0.1887,
|
| 221 |
+
"step": 4200
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.408,
|
| 225 |
+
"grad_norm": 0.11787360906600952,
|
| 226 |
+
"learning_rate": 0.0005566609298217553,
|
| 227 |
+
"loss": 0.187,
|
| 228 |
+
"step": 4400
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.44,
|
| 232 |
+
"eval_loss": 0.2085772007703781,
|
| 233 |
+
"eval_runtime": 27.1615,
|
| 234 |
+
"eval_samples_per_second": 37.222,
|
| 235 |
+
"eval_steps_per_second": 0.589,
|
| 236 |
+
"step": 4500
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.472,
|
| 240 |
+
"grad_norm": 0.09200826287269592,
|
| 241 |
+
"learning_rate": 0.0005229251463209568,
|
| 242 |
+
"loss": 0.1875,
|
| 243 |
+
"step": 4600
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.536,
|
| 247 |
+
"grad_norm": 0.0889899879693985,
|
| 248 |
+
"learning_rate": 0.0004890843320792184,
|
| 249 |
+
"loss": 0.1845,
|
| 250 |
+
"step": 4800
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 1.6,
|
| 254 |
+
"grad_norm": 0.08855367451906204,
|
| 255 |
+
"learning_rate": 0.0004552935275810832,
|
| 256 |
+
"loss": 0.1824,
|
| 257 |
+
"step": 5000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.6,
|
| 261 |
+
"eval_loss": 0.20444034039974213,
|
| 262 |
+
"eval_runtime": 22.73,
|
| 263 |
+
"eval_samples_per_second": 44.479,
|
| 264 |
+
"eval_steps_per_second": 0.704,
|
| 265 |
+
"step": 5000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 1.6640000000000001,
|
| 269 |
+
"grad_norm": 0.09017802029848099,
|
| 270 |
+
"learning_rate": 0.0004217075441932357,
|
| 271 |
+
"loss": 0.1849,
|
| 272 |
+
"step": 5200
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.728,
|
| 276 |
+
"grad_norm": 0.08517798036336899,
|
| 277 |
+
"learning_rate": 0.00038848025490174253,
|
| 278 |
+
"loss": 0.1846,
|
| 279 |
+
"step": 5400
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.76,
|
| 283 |
+
"eval_loss": 0.20072534680366516,
|
| 284 |
+
"eval_runtime": 22.6827,
|
| 285 |
+
"eval_samples_per_second": 44.571,
|
| 286 |
+
"eval_steps_per_second": 0.705,
|
| 287 |
+
"step": 5500
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.792,
|
| 291 |
+
"grad_norm": 0.08856978267431259,
|
| 292 |
+
"learning_rate": 0.00035576388934845007,
|
| 293 |
+
"loss": 0.1821,
|
| 294 |
+
"step": 5600
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.8559999999999999,
|
| 298 |
+
"grad_norm": 0.0986369326710701,
|
| 299 |
+
"learning_rate": 0.00032370833639630423,
|
| 300 |
+
"loss": 0.179,
|
| 301 |
+
"step": 5800
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.92,
|
| 305 |
+
"grad_norm": 0.08553028106689453,
|
| 306 |
+
"learning_rate": 0.00029246045741886697,
|
| 307 |
+
"loss": 0.1797,
|
| 308 |
+
"step": 6000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.92,
|
| 312 |
+
"eval_loss": 0.1978333443403244,
|
| 313 |
+
"eval_runtime": 22.7156,
|
| 314 |
+
"eval_samples_per_second": 44.507,
|
| 315 |
+
"eval_steps_per_second": 0.704,
|
| 316 |
+
"step": 6000
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.984,
|
| 320 |
+
"grad_norm": 0.08506551384925842,
|
| 321 |
+
"learning_rate": 0.00026216341346016616,
|
| 322 |
+
"loss": 0.1796,
|
| 323 |
+
"step": 6200
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 2.048,
|
| 327 |
+
"grad_norm": 0.09524673223495483,
|
| 328 |
+
"learning_rate": 0.00023295600934747395,
|
| 329 |
+
"loss": 0.1607,
|
| 330 |
+
"step": 6400
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 2.08,
|
| 334 |
+
"eval_loss": 0.1975262463092804,
|
| 335 |
+
"eval_runtime": 19.4961,
|
| 336 |
+
"eval_samples_per_second": 51.857,
|
| 337 |
+
"eval_steps_per_second": 0.821,
|
| 338 |
+
"step": 6500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.112,
|
| 342 |
+
"grad_norm": 0.0900162011384964,
|
| 343 |
+
"learning_rate": 0.0002049720577619374,
|
| 344 |
+
"loss": 0.1537,
|
| 345 |
+
"step": 6600
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.176,
|
| 349 |
+
"grad_norm": 0.08766793459653854,
|
| 350 |
+
"learning_rate": 0.00017833976618054676,
|
| 351 |
+
"loss": 0.1544,
|
| 352 |
+
"step": 6800
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 2.24,
|
| 356 |
+
"grad_norm": 0.09870422631502151,
|
| 357 |
+
"learning_rate": 0.0001531811494981501,
|
| 358 |
+
"loss": 0.1522,
|
| 359 |
+
"step": 7000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 2.24,
|
| 363 |
+
"eval_loss": 0.19707150757312775,
|
| 364 |
+
"eval_runtime": 19.4869,
|
| 365 |
+
"eval_samples_per_second": 51.881,
|
| 366 |
+
"eval_steps_per_second": 0.821,
|
| 367 |
+
"step": 7000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 2.304,
|
| 371 |
+
"grad_norm": 0.09113412350416183,
|
| 372 |
+
"learning_rate": 0.0001296114710205592,
|
| 373 |
+
"loss": 0.1539,
|
| 374 |
+
"step": 7200
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.368,
|
| 378 |
+
"grad_norm": 0.09298386424779892,
|
| 379 |
+
"learning_rate": 0.00010773871438982197,
|
| 380 |
+
"loss": 0.1565,
|
| 381 |
+
"step": 7400
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.4,
|
| 385 |
+
"eval_loss": 0.19512927532196045,
|
| 386 |
+
"eval_runtime": 19.4898,
|
| 387 |
+
"eval_samples_per_second": 51.873,
|
| 388 |
+
"eval_steps_per_second": 0.821,
|
| 389 |
+
"step": 7500
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.432,
|
| 393 |
+
"grad_norm": 0.09513990581035614,
|
| 394 |
+
"learning_rate": 8.766308886101404e-05,
|
| 395 |
+
"loss": 0.1551,
|
| 396 |
+
"step": 7600
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 2.496,
|
| 400 |
+
"grad_norm": 0.10141433775424957,
|
| 401 |
+
"learning_rate": 6.947657019710795e-05,
|
| 402 |
+
"loss": 0.1553,
|
| 403 |
+
"step": 7800
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 2.56,
|
| 407 |
+
"grad_norm": 0.09975577890872955,
|
| 408 |
+
"learning_rate": 5.3262479285291874e-05,
|
| 409 |
+
"loss": 0.1533,
|
| 410 |
+
"step": 8000
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.56,
|
| 414 |
+
"eval_loss": 0.19368411600589752,
|
| 415 |
+
"eval_runtime": 19.4935,
|
| 416 |
+
"eval_samples_per_second": 51.864,
|
| 417 |
+
"eval_steps_per_second": 0.821,
|
| 418 |
+
"step": 8000
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 2.624,
|
| 422 |
+
"grad_norm": 0.09386882930994034,
|
| 423 |
+
"learning_rate": 3.90951004052949e-05,
|
| 424 |
+
"loss": 0.1517,
|
| 425 |
+
"step": 8200
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 2.6879999999999997,
|
| 429 |
+
"grad_norm": 0.10441362112760544,
|
| 430 |
+
"learning_rate": 2.703934089860627e-05,
|
| 431 |
+
"loss": 0.1534,
|
| 432 |
+
"step": 8400
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 2.7199999999999998,
|
| 436 |
+
"eval_loss": 0.19298546016216278,
|
| 437 |
+
"eval_runtime": 19.4753,
|
| 438 |
+
"eval_samples_per_second": 51.912,
|
| 439 |
+
"eval_steps_per_second": 0.822,
|
| 440 |
+
"step": 8500
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.752,
|
| 444 |
+
"grad_norm": 0.09755131602287292,
|
| 445 |
+
"learning_rate": 1.715043379780301e-05,
|
| 446 |
+
"loss": 0.1507,
|
| 447 |
+
"step": 8600
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 2.816,
|
| 451 |
+
"grad_norm": 0.1058456152677536,
|
| 452 |
+
"learning_rate": 9.473684778379677e-06,
|
| 453 |
+
"loss": 0.1533,
|
| 454 |
+
"step": 8800
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 2.88,
|
| 458 |
+
"grad_norm": 0.08434511721134186,
|
| 459 |
+
"learning_rate": 4.044264592410641e-06,
|
| 460 |
+
"loss": 0.1521,
|
| 461 |
+
"step": 9000
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.88,
|
| 465 |
+
"eval_loss": 0.1928529292345047,
|
| 466 |
+
"eval_runtime": 19.5071,
|
| 467 |
+
"eval_samples_per_second": 51.827,
|
| 468 |
+
"eval_steps_per_second": 0.82,
|
| 469 |
+
"step": 9000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 2.944,
|
| 473 |
+
"grad_norm": 0.10123102366924286,
|
| 474 |
+
"learning_rate": 8.87047935002272e-07,
|
| 475 |
+
"loss": 0.1495,
|
| 476 |
+
"step": 9200
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 3.0,
|
| 480 |
+
"step": 9375,
|
| 481 |
+
"total_flos": 6.097062592512e+18,
|
| 482 |
+
"train_loss": 0.19507358459472657,
|
| 483 |
+
"train_runtime": 9183.2625,
|
| 484 |
+
"train_samples_per_second": 32.668,
|
| 485 |
+
"train_steps_per_second": 1.021
|
| 486 |
+
}
|
| 487 |
+
],
|
| 488 |
+
"logging_steps": 200,
|
| 489 |
+
"max_steps": 9375,
|
| 490 |
+
"num_input_tokens_seen": 0,
|
| 491 |
+
"num_train_epochs": 3,
|
| 492 |
+
"save_steps": 0,
|
| 493 |
+
"stateful_callbacks": {
|
| 494 |
+
"TrainerControl": {
|
| 495 |
+
"args": {
|
| 496 |
+
"should_epoch_stop": false,
|
| 497 |
+
"should_evaluate": false,
|
| 498 |
+
"should_log": false,
|
| 499 |
+
"should_save": true,
|
| 500 |
+
"should_training_stop": true
|
| 501 |
+
},
|
| 502 |
+
"attributes": {}
|
| 503 |
+
}
|
| 504 |
+
},
|
| 505 |
+
"total_flos": 6.097062592512e+18,
|
| 506 |
+
"train_batch_size": 32,
|
| 507 |
+
"trial_name": null,
|
| 508 |
+
"trial_params": null
|
| 509 |
+
}
|
nl_tasks/exp395/run_ex06/ft/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"T": 1.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"inference_mode": false,
|
| 6 |
+
"layers_to_transform": null,
|
| 7 |
+
"modules_to_save": null,
|
| 8 |
+
"num_rotations": 1,
|
| 9 |
+
"peft_type": "ROTATION",
|
| 10 |
+
"r": 16,
|
| 11 |
+
"revision": null,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"q_proj",
|
| 14 |
+
"v_proj"
|
| 15 |
+
],
|
| 16 |
+
"target_modules_to_skip": null,
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
nl_tasks/exp395/run_ex06/ft/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[PAD]": 32000
|
| 3 |
+
}
|
nl_tasks/exp395/run_ex06/ft/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"pad_token": {
|
| 17 |
+
"content": "[PAD]",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "</s>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
nl_tasks/exp395/run_ex06/ft/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
nl_tasks/exp395/run_ex06/ft/tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
nl_tasks/exp395/run_ex06/ft/tokenizer_config.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": null,
|
| 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 |
+
"32000": {
|
| 31 |
+
"content": "[PAD]",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false,
|
| 36 |
+
"special": true
|
| 37 |
+
}
|
| 38 |
+
},
|
| 39 |
+
"bos_token": "</s>",
|
| 40 |
+
"clean_up_tokenization_spaces": false,
|
| 41 |
+
"eos_token": "</s>",
|
| 42 |
+
"extra_special_tokens": {},
|
| 43 |
+
"legacy": false,
|
| 44 |
+
"model_max_length": 512,
|
| 45 |
+
"pad_token": "[PAD]",
|
| 46 |
+
"padding_side": "right",
|
| 47 |
+
"sp_model_kwargs": {},
|
| 48 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 49 |
+
"unk_token": "</s>",
|
| 50 |
+
"use_default_system_prompt": false
|
| 51 |
+
}
|
nl_tasks/exp395/run_ex06/ft2/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"T": 1.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"layers_to_transform": null,
|
| 7 |
+
"modules_to_save": null,
|
| 8 |
+
"num_rotations": 1,
|
| 9 |
+
"peft_type": "ROTATION",
|
| 10 |
+
"r": 16,
|
| 11 |
+
"revision": null,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"q_proj",
|
| 14 |
+
"v_proj"
|
| 15 |
+
],
|
| 16 |
+
"target_modules_to_skip": null,
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
nl_tasks/exp395/run_ex06/ft2/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:80f122d4725a9564791500cc72f337b41d525cdf466f35d47626f6b9a8b013af
|
| 3 |
+
size 33602915
|
nl_tasks/exp395/run_ex06/trainer_state.json
ADDED
|
@@ -0,0 +1,509 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 9375,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.064,
|
| 14 |
+
"grad_norm": 0.2426149696111679,
|
| 15 |
+
"learning_rate": 0.00199936844222861,
|
| 16 |
+
"loss": 0.3762,
|
| 17 |
+
"step": 200
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.128,
|
| 21 |
+
"grad_norm": 0.23630200326442719,
|
| 22 |
+
"learning_rate": 0.0019946753128925747,
|
| 23 |
+
"loss": 0.3053,
|
| 24 |
+
"step": 400
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.16,
|
| 28 |
+
"eval_loss": 0.2808550000190735,
|
| 29 |
+
"eval_runtime": 19.7751,
|
| 30 |
+
"eval_samples_per_second": 51.125,
|
| 31 |
+
"eval_steps_per_second": 0.809,
|
| 32 |
+
"step": 500
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.192,
|
| 36 |
+
"grad_norm": 0.140605628490448,
|
| 37 |
+
"learning_rate": 0.001985425114055429,
|
| 38 |
+
"loss": 0.2723,
|
| 39 |
+
"step": 600
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.256,
|
| 43 |
+
"grad_norm": 0.12724126875400543,
|
| 44 |
+
"learning_rate": 0.0019716602251735175,
|
| 45 |
+
"loss": 0.2563,
|
| 46 |
+
"step": 800
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.32,
|
| 50 |
+
"grad_norm": 0.09434176236391068,
|
| 51 |
+
"learning_rate": 0.001953443709594746,
|
| 52 |
+
"loss": 0.2481,
|
| 53 |
+
"step": 1000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.32,
|
| 57 |
+
"eval_loss": 0.2517923414707184,
|
| 58 |
+
"eval_runtime": 19.4998,
|
| 59 |
+
"eval_samples_per_second": 51.847,
|
| 60 |
+
"eval_steps_per_second": 0.821,
|
| 61 |
+
"step": 1000
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.384,
|
| 65 |
+
"grad_norm": 0.09533161669969559,
|
| 66 |
+
"learning_rate": 0.0019308590256360988,
|
| 67 |
+
"loss": 0.2449,
|
| 68 |
+
"step": 1200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.448,
|
| 72 |
+
"grad_norm": 0.09580417722463608,
|
| 73 |
+
"learning_rate": 0.0019040096442223358,
|
| 74 |
+
"loss": 0.2388,
|
| 75 |
+
"step": 1400
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.48,
|
| 79 |
+
"eval_loss": 0.24148738384246826,
|
| 80 |
+
"eval_runtime": 19.5462,
|
| 81 |
+
"eval_samples_per_second": 51.724,
|
| 82 |
+
"eval_steps_per_second": 0.819,
|
| 83 |
+
"step": 1500
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.512,
|
| 87 |
+
"grad_norm": 0.08012712001800537,
|
| 88 |
+
"learning_rate": 0.0018730185748376353,
|
| 89 |
+
"loss": 0.2334,
|
| 90 |
+
"step": 1600
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.576,
|
| 94 |
+
"grad_norm": 0.08019606024026871,
|
| 95 |
+
"learning_rate": 0.0018380278019620281,
|
| 96 |
+
"loss": 0.2292,
|
| 97 |
+
"step": 1800
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.64,
|
| 101 |
+
"grad_norm": 0.0736006572842598,
|
| 102 |
+
"learning_rate": 0.0017991976345745596,
|
| 103 |
+
"loss": 0.2276,
|
| 104 |
+
"step": 2000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.64,
|
| 108 |
+
"eval_loss": 0.22908221185207367,
|
| 109 |
+
"eval_runtime": 19.5153,
|
| 110 |
+
"eval_samples_per_second": 51.805,
|
| 111 |
+
"eval_steps_per_second": 0.82,
|
| 112 |
+
"step": 2000
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.704,
|
| 116 |
+
"grad_norm": 0.06510240584611893,
|
| 117 |
+
"learning_rate": 0.0017567059717034153,
|
| 118 |
+
"loss": 0.2239,
|
| 119 |
+
"step": 2200
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.768,
|
| 123 |
+
"grad_norm": 0.06003573164343834,
|
| 124 |
+
"learning_rate": 0.0017107474873878648,
|
| 125 |
+
"loss": 0.2219,
|
| 126 |
+
"step": 2400
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.8,
|
| 130 |
+
"eval_loss": 0.22465121746063232,
|
| 131 |
+
"eval_runtime": 19.4747,
|
| 132 |
+
"eval_samples_per_second": 51.914,
|
| 133 |
+
"eval_steps_per_second": 0.822,
|
| 134 |
+
"step": 2500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.832,
|
| 138 |
+
"grad_norm": 0.07121224701404572,
|
| 139 |
+
"learning_rate": 0.0016615327387860851,
|
| 140 |
+
"loss": 0.2197,
|
| 141 |
+
"step": 2600
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.896,
|
| 145 |
+
"grad_norm": 0.06296424567699432,
|
| 146 |
+
"learning_rate": 0.001609287201515044,
|
| 147 |
+
"loss": 0.2191,
|
| 148 |
+
"step": 2800
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.96,
|
| 152 |
+
"grad_norm": 0.05874966084957123,
|
| 153 |
+
"learning_rate": 0.0015542502366419986,
|
| 154 |
+
"loss": 0.2121,
|
| 155 |
+
"step": 3000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.96,
|
| 159 |
+
"eval_loss": 0.2207345813512802,
|
| 160 |
+
"eval_runtime": 19.5619,
|
| 161 |
+
"eval_samples_per_second": 51.682,
|
| 162 |
+
"eval_steps_per_second": 0.818,
|
| 163 |
+
"step": 3000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.024,
|
| 167 |
+
"grad_norm": 0.0780828595161438,
|
| 168 |
+
"learning_rate": 0.001496673994060291,
|
| 169 |
+
"loss": 0.2069,
|
| 170 |
+
"step": 3200
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.088,
|
| 174 |
+
"grad_norm": 0.06145294010639191,
|
| 175 |
+
"learning_rate": 0.0014368222572736001,
|
| 176 |
+
"loss": 0.1902,
|
| 177 |
+
"step": 3400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.12,
|
| 181 |
+
"eval_loss": 0.21642106771469116,
|
| 182 |
+
"eval_runtime": 19.4895,
|
| 183 |
+
"eval_samples_per_second": 51.874,
|
| 184 |
+
"eval_steps_per_second": 0.821,
|
| 185 |
+
"step": 3500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.152,
|
| 189 |
+
"grad_norm": 0.0599561408162117,
|
| 190 |
+
"learning_rate": 0.0013749692348812186,
|
| 191 |
+
"loss": 0.1907,
|
| 192 |
+
"step": 3600
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.216,
|
| 196 |
+
"grad_norm": 0.060904331505298615,
|
| 197 |
+
"learning_rate": 0.0013113983043011267,
|
| 198 |
+
"loss": 0.1905,
|
| 199 |
+
"step": 3800
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.28,
|
| 203 |
+
"grad_norm": 0.05701998248696327,
|
| 204 |
+
"learning_rate": 0.0012464007134864482,
|
| 205 |
+
"loss": 0.1885,
|
| 206 |
+
"step": 4000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.28,
|
| 210 |
+
"eval_loss": 0.2138749361038208,
|
| 211 |
+
"eval_runtime": 19.5319,
|
| 212 |
+
"eval_samples_per_second": 51.761,
|
| 213 |
+
"eval_steps_per_second": 0.819,
|
| 214 |
+
"step": 4000
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.3439999999999999,
|
| 218 |
+
"grad_norm": 0.051733970642089844,
|
| 219 |
+
"learning_rate": 0.0011802742465833352,
|
| 220 |
+
"loss": 0.1902,
|
| 221 |
+
"step": 4200
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.408,
|
| 225 |
+
"grad_norm": 0.07325556129217148,
|
| 226 |
+
"learning_rate": 0.0011133218596435107,
|
| 227 |
+
"loss": 0.1884,
|
| 228 |
+
"step": 4400
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.44,
|
| 232 |
+
"eval_loss": 0.20981121063232422,
|
| 233 |
+
"eval_runtime": 19.5236,
|
| 234 |
+
"eval_samples_per_second": 51.784,
|
| 235 |
+
"eval_steps_per_second": 0.82,
|
| 236 |
+
"step": 4500
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.472,
|
| 240 |
+
"grad_norm": 0.05644873157143593,
|
| 241 |
+
"learning_rate": 0.0010458502926419136,
|
| 242 |
+
"loss": 0.189,
|
| 243 |
+
"step": 4600
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.536,
|
| 247 |
+
"grad_norm": 0.07849732786417007,
|
| 248 |
+
"learning_rate": 0.0009781686641584368,
|
| 249 |
+
"loss": 0.1864,
|
| 250 |
+
"step": 4800
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 1.6,
|
| 254 |
+
"grad_norm": 0.05307694151997566,
|
| 255 |
+
"learning_rate": 0.0009105870551621664,
|
| 256 |
+
"loss": 0.1842,
|
| 257 |
+
"step": 5000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.6,
|
| 261 |
+
"eval_loss": 0.20523199439048767,
|
| 262 |
+
"eval_runtime": 19.534,
|
| 263 |
+
"eval_samples_per_second": 51.756,
|
| 264 |
+
"eval_steps_per_second": 0.819,
|
| 265 |
+
"step": 5000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 1.6640000000000001,
|
| 269 |
+
"grad_norm": 0.04950845614075661,
|
| 270 |
+
"learning_rate": 0.0008434150883864713,
|
| 271 |
+
"loss": 0.1866,
|
| 272 |
+
"step": 5200
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.728,
|
| 276 |
+
"grad_norm": 0.05008252337574959,
|
| 277 |
+
"learning_rate": 0.0007769605098034851,
|
| 278 |
+
"loss": 0.1859,
|
| 279 |
+
"step": 5400
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.76,
|
| 283 |
+
"eval_loss": 0.20152263343334198,
|
| 284 |
+
"eval_runtime": 19.4908,
|
| 285 |
+
"eval_samples_per_second": 51.871,
|
| 286 |
+
"eval_steps_per_second": 0.821,
|
| 287 |
+
"step": 5500
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.792,
|
| 291 |
+
"grad_norm": 0.04900854080915451,
|
| 292 |
+
"learning_rate": 0.0007115277786969001,
|
| 293 |
+
"loss": 0.1831,
|
| 294 |
+
"step": 5600
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.8559999999999999,
|
| 298 |
+
"grad_norm": 0.058961208909749985,
|
| 299 |
+
"learning_rate": 0.0006474166727926085,
|
| 300 |
+
"loss": 0.1805,
|
| 301 |
+
"step": 5800
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.92,
|
| 305 |
+
"grad_norm": 0.05107206851243973,
|
| 306 |
+
"learning_rate": 0.0005849209148377339,
|
| 307 |
+
"loss": 0.181,
|
| 308 |
+
"step": 6000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.92,
|
| 312 |
+
"eval_loss": 0.19884061813354492,
|
| 313 |
+
"eval_runtime": 19.5072,
|
| 314 |
+
"eval_samples_per_second": 51.827,
|
| 315 |
+
"eval_steps_per_second": 0.82,
|
| 316 |
+
"step": 6000
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.984,
|
| 320 |
+
"grad_norm": 0.048386573791503906,
|
| 321 |
+
"learning_rate": 0.0005243268269203323,
|
| 322 |
+
"loss": 0.1806,
|
| 323 |
+
"step": 6200
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 2.048,
|
| 327 |
+
"grad_norm": 0.055718015879392624,
|
| 328 |
+
"learning_rate": 0.0004659120186949479,
|
| 329 |
+
"loss": 0.1617,
|
| 330 |
+
"step": 6400
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 2.08,
|
| 334 |
+
"eval_loss": 0.19804580509662628,
|
| 335 |
+
"eval_runtime": 19.5582,
|
| 336 |
+
"eval_samples_per_second": 51.692,
|
| 337 |
+
"eval_steps_per_second": 0.818,
|
| 338 |
+
"step": 6500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.112,
|
| 342 |
+
"grad_norm": 0.05408688262104988,
|
| 343 |
+
"learning_rate": 0.0004099441155238748,
|
| 344 |
+
"loss": 0.1543,
|
| 345 |
+
"step": 6600
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.176,
|
| 349 |
+
"grad_norm": 0.05257072672247887,
|
| 350 |
+
"learning_rate": 0.0003566795323610935,
|
| 351 |
+
"loss": 0.1549,
|
| 352 |
+
"step": 6800
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 2.24,
|
| 356 |
+
"grad_norm": 0.05060563609004021,
|
| 357 |
+
"learning_rate": 0.0003063622989963002,
|
| 358 |
+
"loss": 0.1528,
|
| 359 |
+
"step": 7000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 2.24,
|
| 363 |
+
"eval_loss": 0.19734542071819305,
|
| 364 |
+
"eval_runtime": 19.5386,
|
| 365 |
+
"eval_samples_per_second": 51.744,
|
| 366 |
+
"eval_steps_per_second": 0.819,
|
| 367 |
+
"step": 7000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 2.304,
|
| 371 |
+
"grad_norm": 0.05511431023478508,
|
| 372 |
+
"learning_rate": 0.0002592229420411184,
|
| 373 |
+
"loss": 0.1543,
|
| 374 |
+
"step": 7200
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.368,
|
| 378 |
+
"grad_norm": 0.07401052117347717,
|
| 379 |
+
"learning_rate": 0.00021547742877964395,
|
| 380 |
+
"loss": 0.1567,
|
| 381 |
+
"step": 7400
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.4,
|
| 385 |
+
"eval_loss": 0.1949290782213211,
|
| 386 |
+
"eval_runtime": 19.4974,
|
| 387 |
+
"eval_samples_per_second": 51.853,
|
| 388 |
+
"eval_steps_per_second": 0.821,
|
| 389 |
+
"step": 7500
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.432,
|
| 393 |
+
"grad_norm": 0.05771145224571228,
|
| 394 |
+
"learning_rate": 0.00017532617772202807,
|
| 395 |
+
"loss": 0.1557,
|
| 396 |
+
"step": 7600
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 2.496,
|
| 400 |
+
"grad_norm": 0.05626155808568001,
|
| 401 |
+
"learning_rate": 0.0001389531403942159,
|
| 402 |
+
"loss": 0.1559,
|
| 403 |
+
"step": 7800
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 2.56,
|
| 407 |
+
"grad_norm": 0.05681011080741882,
|
| 408 |
+
"learning_rate": 0.00010652495857058375,
|
| 409 |
+
"loss": 0.1536,
|
| 410 |
+
"step": 8000
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.56,
|
| 414 |
+
"eval_loss": 0.19341105222702026,
|
| 415 |
+
"eval_runtime": 33.6538,
|
| 416 |
+
"eval_samples_per_second": 30.041,
|
| 417 |
+
"eval_steps_per_second": 0.475,
|
| 418 |
+
"step": 8000
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 2.624,
|
| 422 |
+
"grad_norm": 0.05382503569126129,
|
| 423 |
+
"learning_rate": 7.81902008105898e-05,
|
| 424 |
+
"loss": 0.1516,
|
| 425 |
+
"step": 8200
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 2.6879999999999997,
|
| 429 |
+
"grad_norm": 0.06113835424184799,
|
| 430 |
+
"learning_rate": 5.407868179721254e-05,
|
| 431 |
+
"loss": 0.1535,
|
| 432 |
+
"step": 8400
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 2.7199999999999998,
|
| 436 |
+
"eval_loss": 0.1927892565727234,
|
| 437 |
+
"eval_runtime": 19.4973,
|
| 438 |
+
"eval_samples_per_second": 51.853,
|
| 439 |
+
"eval_steps_per_second": 0.821,
|
| 440 |
+
"step": 8500
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.752,
|
| 444 |
+
"grad_norm": 0.055167559534311295,
|
| 445 |
+
"learning_rate": 3.430086759560602e-05,
|
| 446 |
+
"loss": 0.1506,
|
| 447 |
+
"step": 8600
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 2.816,
|
| 451 |
+
"grad_norm": 0.06194595992565155,
|
| 452 |
+
"learning_rate": 1.8947369556759353e-05,
|
| 453 |
+
"loss": 0.1537,
|
| 454 |
+
"step": 8800
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 2.88,
|
| 458 |
+
"grad_norm": 0.0478057824075222,
|
| 459 |
+
"learning_rate": 8.088529184821281e-06,
|
| 460 |
+
"loss": 0.1522,
|
| 461 |
+
"step": 9000
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.88,
|
| 465 |
+
"eval_loss": 0.1926267296075821,
|
| 466 |
+
"eval_runtime": 25.1004,
|
| 467 |
+
"eval_samples_per_second": 40.278,
|
| 468 |
+
"eval_steps_per_second": 0.637,
|
| 469 |
+
"step": 9000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 2.944,
|
| 473 |
+
"grad_norm": 0.06033524498343468,
|
| 474 |
+
"learning_rate": 1.774095870004544e-06,
|
| 475 |
+
"loss": 0.1493,
|
| 476 |
+
"step": 9200
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 3.0,
|
| 480 |
+
"step": 9375,
|
| 481 |
+
"total_flos": 6.097062592512e+18,
|
| 482 |
+
"train_loss": 0.19569317911783854,
|
| 483 |
+
"train_runtime": 9108.3132,
|
| 484 |
+
"train_samples_per_second": 32.937,
|
| 485 |
+
"train_steps_per_second": 1.029
|
| 486 |
+
}
|
| 487 |
+
],
|
| 488 |
+
"logging_steps": 200,
|
| 489 |
+
"max_steps": 9375,
|
| 490 |
+
"num_input_tokens_seen": 0,
|
| 491 |
+
"num_train_epochs": 3,
|
| 492 |
+
"save_steps": 0,
|
| 493 |
+
"stateful_callbacks": {
|
| 494 |
+
"TrainerControl": {
|
| 495 |
+
"args": {
|
| 496 |
+
"should_epoch_stop": false,
|
| 497 |
+
"should_evaluate": false,
|
| 498 |
+
"should_log": false,
|
| 499 |
+
"should_save": true,
|
| 500 |
+
"should_training_stop": true
|
| 501 |
+
},
|
| 502 |
+
"attributes": {}
|
| 503 |
+
}
|
| 504 |
+
},
|
| 505 |
+
"total_flos": 6.097062592512e+18,
|
| 506 |
+
"train_batch_size": 32,
|
| 507 |
+
"trial_name": null,
|
| 508 |
+
"trial_params": null
|
| 509 |
+
}
|
nl_tasks/exp395/run_ex07/ft/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"T": 1.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"inference_mode": false,
|
| 6 |
+
"layers_to_transform": null,
|
| 7 |
+
"modules_to_save": null,
|
| 8 |
+
"num_rotations": 1,
|
| 9 |
+
"peft_type": "ROTATION",
|
| 10 |
+
"r": 16,
|
| 11 |
+
"revision": null,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"v_proj",
|
| 14 |
+
"q_proj"
|
| 15 |
+
],
|
| 16 |
+
"target_modules_to_skip": null,
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
nl_tasks/exp395/run_ex07/ft/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[PAD]": 32000
|
| 3 |
+
}
|
nl_tasks/exp395/run_ex07/ft/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"pad_token": {
|
| 17 |
+
"content": "[PAD]",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "</s>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
nl_tasks/exp395/run_ex07/ft/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
nl_tasks/exp395/run_ex07/ft/tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
nl_tasks/exp395/run_ex07/ft/tokenizer_config.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": null,
|
| 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 |
+
"32000": {
|
| 31 |
+
"content": "[PAD]",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false,
|
| 36 |
+
"special": true
|
| 37 |
+
}
|
| 38 |
+
},
|
| 39 |
+
"bos_token": "</s>",
|
| 40 |
+
"clean_up_tokenization_spaces": false,
|
| 41 |
+
"eos_token": "</s>",
|
| 42 |
+
"extra_special_tokens": {},
|
| 43 |
+
"legacy": false,
|
| 44 |
+
"model_max_length": 512,
|
| 45 |
+
"pad_token": "[PAD]",
|
| 46 |
+
"padding_side": "right",
|
| 47 |
+
"sp_model_kwargs": {},
|
| 48 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 49 |
+
"unk_token": "</s>",
|
| 50 |
+
"use_default_system_prompt": false
|
| 51 |
+
}
|
nl_tasks/exp395/run_ex07/ft2/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"T": 1.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"layers_to_transform": null,
|
| 7 |
+
"modules_to_save": null,
|
| 8 |
+
"num_rotations": 1,
|
| 9 |
+
"peft_type": "ROTATION",
|
| 10 |
+
"r": 16,
|
| 11 |
+
"revision": null,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"v_proj",
|
| 14 |
+
"q_proj"
|
| 15 |
+
],
|
| 16 |
+
"target_modules_to_skip": null,
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
nl_tasks/exp395/run_ex07/ft2/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cfc792c0299154bc4d052cdf373ee5d1328a23a53804f4af7642a0da515cc656
|
| 3 |
+
size 33602915
|
nl_tasks/exp395/run_ex07/trainer_state.json
ADDED
|
@@ -0,0 +1,509 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 9375,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.064,
|
| 14 |
+
"grad_norm": 0.17820023000240326,
|
| 15 |
+
"learning_rate": 0.004998421105571525,
|
| 16 |
+
"loss": 0.4046,
|
| 17 |
+
"step": 200
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.128,
|
| 21 |
+
"grad_norm": 0.11274894326925278,
|
| 22 |
+
"learning_rate": 0.0049866882822314365,
|
| 23 |
+
"loss": 0.294,
|
| 24 |
+
"step": 400
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.16,
|
| 28 |
+
"eval_loss": 0.27918341755867004,
|
| 29 |
+
"eval_runtime": 28.1591,
|
| 30 |
+
"eval_samples_per_second": 35.903,
|
| 31 |
+
"eval_steps_per_second": 0.568,
|
| 32 |
+
"step": 500
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.192,
|
| 36 |
+
"grad_norm": 0.07643634080886841,
|
| 37 |
+
"learning_rate": 0.004963562785138573,
|
| 38 |
+
"loss": 0.2697,
|
| 39 |
+
"step": 600
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.256,
|
| 43 |
+
"grad_norm": 0.07283908873796463,
|
| 44 |
+
"learning_rate": 0.004929150562933794,
|
| 45 |
+
"loss": 0.2571,
|
| 46 |
+
"step": 800
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.32,
|
| 50 |
+
"grad_norm": 0.06430114805698395,
|
| 51 |
+
"learning_rate": 0.004883609273986864,
|
| 52 |
+
"loss": 0.2491,
|
| 53 |
+
"step": 1000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.32,
|
| 57 |
+
"eval_loss": 0.2539050281047821,
|
| 58 |
+
"eval_runtime": 22.7763,
|
| 59 |
+
"eval_samples_per_second": 44.388,
|
| 60 |
+
"eval_steps_per_second": 0.702,
|
| 61 |
+
"step": 1000
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.384,
|
| 65 |
+
"grad_norm": 0.0629628598690033,
|
| 66 |
+
"learning_rate": 0.004827147564090247,
|
| 67 |
+
"loss": 0.2473,
|
| 68 |
+
"step": 1200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.448,
|
| 72 |
+
"grad_norm": 0.05335867032408714,
|
| 73 |
+
"learning_rate": 0.0047600241105558395,
|
| 74 |
+
"loss": 0.2409,
|
| 75 |
+
"step": 1400
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.48,
|
| 79 |
+
"eval_loss": 0.2420750856399536,
|
| 80 |
+
"eval_runtime": 22.8179,
|
| 81 |
+
"eval_samples_per_second": 44.307,
|
| 82 |
+
"eval_steps_per_second": 0.701,
|
| 83 |
+
"step": 1500
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.512,
|
| 87 |
+
"grad_norm": 0.04761116951704025,
|
| 88 |
+
"learning_rate": 0.0046825464370940885,
|
| 89 |
+
"loss": 0.2349,
|
| 90 |
+
"step": 1600
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.576,
|
| 94 |
+
"grad_norm": 0.043634187430143356,
|
| 95 |
+
"learning_rate": 0.00459506950490507,
|
| 96 |
+
"loss": 0.2322,
|
| 97 |
+
"step": 1800
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.64,
|
| 101 |
+
"grad_norm": 0.04088456556200981,
|
| 102 |
+
"learning_rate": 0.004497994086436399,
|
| 103 |
+
"loss": 0.2302,
|
| 104 |
+
"step": 2000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.64,
|
| 108 |
+
"eval_loss": 0.23234312236309052,
|
| 109 |
+
"eval_runtime": 22.7709,
|
| 110 |
+
"eval_samples_per_second": 44.399,
|
| 111 |
+
"eval_steps_per_second": 0.703,
|
| 112 |
+
"step": 2000
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.704,
|
| 116 |
+
"grad_norm": 0.03686574473977089,
|
| 117 |
+
"learning_rate": 0.004391764929258538,
|
| 118 |
+
"loss": 0.2264,
|
| 119 |
+
"step": 2200
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.768,
|
| 123 |
+
"grad_norm": 0.033889204263687134,
|
| 124 |
+
"learning_rate": 0.004276868718469662,
|
| 125 |
+
"loss": 0.2246,
|
| 126 |
+
"step": 2400
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.8,
|
| 130 |
+
"eval_loss": 0.22624295949935913,
|
| 131 |
+
"eval_runtime": 19.4841,
|
| 132 |
+
"eval_samples_per_second": 51.888,
|
| 133 |
+
"eval_steps_per_second": 0.821,
|
| 134 |
+
"step": 2500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.832,
|
| 138 |
+
"grad_norm": 0.04181356728076935,
|
| 139 |
+
"learning_rate": 0.0041538318469652126,
|
| 140 |
+
"loss": 0.2222,
|
| 141 |
+
"step": 2600
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.896,
|
| 145 |
+
"grad_norm": 0.03593447804450989,
|
| 146 |
+
"learning_rate": 0.00402321800378761,
|
| 147 |
+
"loss": 0.2219,
|
| 148 |
+
"step": 2800
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.96,
|
| 152 |
+
"grad_norm": 0.031546611338853836,
|
| 153 |
+
"learning_rate": 0.0038856255916049965,
|
| 154 |
+
"loss": 0.2149,
|
| 155 |
+
"step": 3000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.96,
|
| 159 |
+
"eval_loss": 0.22304701805114746,
|
| 160 |
+
"eval_runtime": 27.9954,
|
| 161 |
+
"eval_samples_per_second": 36.113,
|
| 162 |
+
"eval_steps_per_second": 0.572,
|
| 163 |
+
"step": 3000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.024,
|
| 167 |
+
"grad_norm": 0.037926312536001205,
|
| 168 |
+
"learning_rate": 0.0037416849851507274,
|
| 169 |
+
"loss": 0.2094,
|
| 170 |
+
"step": 3200
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.088,
|
| 174 |
+
"grad_norm": 0.030403969809412956,
|
| 175 |
+
"learning_rate": 0.0035920556431840002,
|
| 176 |
+
"loss": 0.1939,
|
| 177 |
+
"step": 3400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.12,
|
| 181 |
+
"eval_loss": 0.21854709088802338,
|
| 182 |
+
"eval_runtime": 19.4946,
|
| 183 |
+
"eval_samples_per_second": 51.86,
|
| 184 |
+
"eval_steps_per_second": 0.821,
|
| 185 |
+
"step": 3500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.152,
|
| 189 |
+
"grad_norm": 0.03306721895933151,
|
| 190 |
+
"learning_rate": 0.0034374230872030466,
|
| 191 |
+
"loss": 0.1944,
|
| 192 |
+
"step": 3600
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.216,
|
| 196 |
+
"grad_norm": 0.03182404488325119,
|
| 197 |
+
"learning_rate": 0.0032784957607528164,
|
| 198 |
+
"loss": 0.1948,
|
| 199 |
+
"step": 3800
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.28,
|
| 203 |
+
"grad_norm": 0.025607936084270477,
|
| 204 |
+
"learning_rate": 0.003116001783716121,
|
| 205 |
+
"loss": 0.1929,
|
| 206 |
+
"step": 4000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.28,
|
| 210 |
+
"eval_loss": 0.2149907499551773,
|
| 211 |
+
"eval_runtime": 25.792,
|
| 212 |
+
"eval_samples_per_second": 39.198,
|
| 213 |
+
"eval_steps_per_second": 0.62,
|
| 214 |
+
"step": 4000
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.3439999999999999,
|
| 218 |
+
"grad_norm": 0.028963735327124596,
|
| 219 |
+
"learning_rate": 0.0029506856164583377,
|
| 220 |
+
"loss": 0.1944,
|
| 221 |
+
"step": 4200
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.408,
|
| 225 |
+
"grad_norm": 0.03618050366640091,
|
| 226 |
+
"learning_rate": 0.0027833046491087766,
|
| 227 |
+
"loss": 0.1928,
|
| 228 |
+
"step": 4400
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.44,
|
| 232 |
+
"eval_loss": 0.21210040152072906,
|
| 233 |
+
"eval_runtime": 19.4863,
|
| 234 |
+
"eval_samples_per_second": 51.882,
|
| 235 |
+
"eval_steps_per_second": 0.821,
|
| 236 |
+
"step": 4500
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.472,
|
| 240 |
+
"grad_norm": 0.03128841519355774,
|
| 241 |
+
"learning_rate": 0.002614625731604784,
|
| 242 |
+
"loss": 0.1935,
|
| 243 |
+
"step": 4600
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.536,
|
| 247 |
+
"grad_norm": 0.026794707402586937,
|
| 248 |
+
"learning_rate": 0.002445421660396092,
|
| 249 |
+
"loss": 0.1905,
|
| 250 |
+
"step": 4800
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 1.6,
|
| 254 |
+
"grad_norm": 0.028422104194760323,
|
| 255 |
+
"learning_rate": 0.0022764676379054163,
|
| 256 |
+
"loss": 0.1878,
|
| 257 |
+
"step": 5000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.6,
|
| 261 |
+
"eval_loss": 0.20722173154354095,
|
| 262 |
+
"eval_runtime": 19.4745,
|
| 263 |
+
"eval_samples_per_second": 51.914,
|
| 264 |
+
"eval_steps_per_second": 0.822,
|
| 265 |
+
"step": 5000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 1.6640000000000001,
|
| 269 |
+
"grad_norm": 0.02493426389992237,
|
| 270 |
+
"learning_rate": 0.0021085377209661784,
|
| 271 |
+
"loss": 0.1904,
|
| 272 |
+
"step": 5200
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.728,
|
| 276 |
+
"grad_norm": 0.026702167466282845,
|
| 277 |
+
"learning_rate": 0.0019424012745087127,
|
| 278 |
+
"loss": 0.19,
|
| 279 |
+
"step": 5400
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.76,
|
| 283 |
+
"eval_loss": 0.20289908349514008,
|
| 284 |
+
"eval_runtime": 26.7848,
|
| 285 |
+
"eval_samples_per_second": 37.745,
|
| 286 |
+
"eval_steps_per_second": 0.597,
|
| 287 |
+
"step": 5500
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.792,
|
| 291 |
+
"grad_norm": 0.026274586096405983,
|
| 292 |
+
"learning_rate": 0.0017788194467422502,
|
| 293 |
+
"loss": 0.187,
|
| 294 |
+
"step": 5600
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.8559999999999999,
|
| 298 |
+
"grad_norm": 0.02791573852300644,
|
| 299 |
+
"learning_rate": 0.001618541681981521,
|
| 300 |
+
"loss": 0.1837,
|
| 301 |
+
"step": 5800
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.92,
|
| 305 |
+
"grad_norm": 0.024751191958785057,
|
| 306 |
+
"learning_rate": 0.0014623022870943348,
|
| 307 |
+
"loss": 0.1841,
|
| 308 |
+
"step": 6000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.92,
|
| 312 |
+
"eval_loss": 0.19980530440807343,
|
| 313 |
+
"eval_runtime": 19.4905,
|
| 314 |
+
"eval_samples_per_second": 51.871,
|
| 315 |
+
"eval_steps_per_second": 0.821,
|
| 316 |
+
"step": 6000
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.984,
|
| 320 |
+
"grad_norm": 0.027469471096992493,
|
| 321 |
+
"learning_rate": 0.0013108170673008306,
|
| 322 |
+
"loss": 0.1838,
|
| 323 |
+
"step": 6200
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 2.048,
|
| 327 |
+
"grad_norm": 0.028063790872693062,
|
| 328 |
+
"learning_rate": 0.0011647800467373699,
|
| 329 |
+
"loss": 0.1639,
|
| 330 |
+
"step": 6400
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 2.08,
|
| 334 |
+
"eval_loss": 0.19846861064434052,
|
| 335 |
+
"eval_runtime": 28.4577,
|
| 336 |
+
"eval_samples_per_second": 35.526,
|
| 337 |
+
"eval_steps_per_second": 0.562,
|
| 338 |
+
"step": 6500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.112,
|
| 342 |
+
"grad_norm": 0.02692621760070324,
|
| 343 |
+
"learning_rate": 0.001024860288809687,
|
| 344 |
+
"loss": 0.1568,
|
| 345 |
+
"step": 6600
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.176,
|
| 349 |
+
"grad_norm": 0.02786995656788349,
|
| 350 |
+
"learning_rate": 0.0008916988309027338,
|
| 351 |
+
"loss": 0.1571,
|
| 352 |
+
"step": 6800
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 2.24,
|
| 356 |
+
"grad_norm": 0.02808111160993576,
|
| 357 |
+
"learning_rate": 0.0007659057474907505,
|
| 358 |
+
"loss": 0.155,
|
| 359 |
+
"step": 7000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 2.24,
|
| 363 |
+
"eval_loss": 0.1971956193447113,
|
| 364 |
+
"eval_runtime": 19.5072,
|
| 365 |
+
"eval_samples_per_second": 51.827,
|
| 366 |
+
"eval_steps_per_second": 0.82,
|
| 367 |
+
"step": 7000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 2.304,
|
| 371 |
+
"grad_norm": 0.02999687008559704,
|
| 372 |
+
"learning_rate": 0.000648057355102796,
|
| 373 |
+
"loss": 0.1566,
|
| 374 |
+
"step": 7200
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.368,
|
| 378 |
+
"grad_norm": 0.027220070362091064,
|
| 379 |
+
"learning_rate": 0.0005386935719491098,
|
| 380 |
+
"loss": 0.1587,
|
| 381 |
+
"step": 7400
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.4,
|
| 385 |
+
"eval_loss": 0.19491882622241974,
|
| 386 |
+
"eval_runtime": 26.4649,
|
| 387 |
+
"eval_samples_per_second": 38.202,
|
| 388 |
+
"eval_steps_per_second": 0.605,
|
| 389 |
+
"step": 7500
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.432,
|
| 393 |
+
"grad_norm": 0.030793363228440285,
|
| 394 |
+
"learning_rate": 0.0004383154443050702,
|
| 395 |
+
"loss": 0.1573,
|
| 396 |
+
"step": 7600
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 2.496,
|
| 400 |
+
"grad_norm": 0.03137922286987305,
|
| 401 |
+
"learning_rate": 0.00034738285098553977,
|
| 402 |
+
"loss": 0.1573,
|
| 403 |
+
"step": 7800
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 2.56,
|
| 407 |
+
"grad_norm": 0.0298843365162611,
|
| 408 |
+
"learning_rate": 0.00026631239642645936,
|
| 409 |
+
"loss": 0.155,
|
| 410 |
+
"step": 8000
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.56,
|
| 414 |
+
"eval_loss": 0.1927516758441925,
|
| 415 |
+
"eval_runtime": 26.2373,
|
| 416 |
+
"eval_samples_per_second": 38.533,
|
| 417 |
+
"eval_steps_per_second": 0.61,
|
| 418 |
+
"step": 8000
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 2.624,
|
| 422 |
+
"grad_norm": 0.0275055430829525,
|
| 423 |
+
"learning_rate": 0.0001954755020264745,
|
| 424 |
+
"loss": 0.1531,
|
| 425 |
+
"step": 8200
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 2.6879999999999997,
|
| 429 |
+
"grad_norm": 0.02978612296283245,
|
| 430 |
+
"learning_rate": 0.00013519670449303135,
|
| 431 |
+
"loss": 0.1549,
|
| 432 |
+
"step": 8400
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 2.7199999999999998,
|
| 436 |
+
"eval_loss": 0.19178353250026703,
|
| 437 |
+
"eval_runtime": 26.5213,
|
| 438 |
+
"eval_samples_per_second": 38.12,
|
| 439 |
+
"eval_steps_per_second": 0.603,
|
| 440 |
+
"step": 8500
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.752,
|
| 444 |
+
"grad_norm": 0.028374744579195976,
|
| 445 |
+
"learning_rate": 8.575216898901506e-05,
|
| 446 |
+
"loss": 0.1517,
|
| 447 |
+
"step": 8600
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 2.816,
|
| 451 |
+
"grad_norm": 0.03000253438949585,
|
| 452 |
+
"learning_rate": 4.736842389189838e-05,
|
| 453 |
+
"loss": 0.1544,
|
| 454 |
+
"step": 8800
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 2.88,
|
| 458 |
+
"grad_norm": 0.025454407557845116,
|
| 459 |
+
"learning_rate": 2.02213229620532e-05,
|
| 460 |
+
"loss": 0.1533,
|
| 461 |
+
"step": 9000
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.88,
|
| 465 |
+
"eval_loss": 0.19136634469032288,
|
| 466 |
+
"eval_runtime": 19.5667,
|
| 467 |
+
"eval_samples_per_second": 51.67,
|
| 468 |
+
"eval_steps_per_second": 0.818,
|
| 469 |
+
"step": 9000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 2.944,
|
| 473 |
+
"grad_norm": 0.030465198680758476,
|
| 474 |
+
"learning_rate": 4.43523967501136e-06,
|
| 475 |
+
"loss": 0.1502,
|
| 476 |
+
"step": 9200
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 3.0,
|
| 480 |
+
"step": 9375,
|
| 481 |
+
"total_flos": 6.097062592512e+18,
|
| 482 |
+
"train_loss": 0.1983985675048828,
|
| 483 |
+
"train_runtime": 10216.7969,
|
| 484 |
+
"train_samples_per_second": 29.363,
|
| 485 |
+
"train_steps_per_second": 0.918
|
| 486 |
+
}
|
| 487 |
+
],
|
| 488 |
+
"logging_steps": 200,
|
| 489 |
+
"max_steps": 9375,
|
| 490 |
+
"num_input_tokens_seen": 0,
|
| 491 |
+
"num_train_epochs": 3,
|
| 492 |
+
"save_steps": 0,
|
| 493 |
+
"stateful_callbacks": {
|
| 494 |
+
"TrainerControl": {
|
| 495 |
+
"args": {
|
| 496 |
+
"should_epoch_stop": false,
|
| 497 |
+
"should_evaluate": false,
|
| 498 |
+
"should_log": false,
|
| 499 |
+
"should_save": true,
|
| 500 |
+
"should_training_stop": true
|
| 501 |
+
},
|
| 502 |
+
"attributes": {}
|
| 503 |
+
}
|
| 504 |
+
},
|
| 505 |
+
"total_flos": 6.097062592512e+18,
|
| 506 |
+
"train_batch_size": 32,
|
| 507 |
+
"trial_name": null,
|
| 508 |
+
"trial_params": null
|
| 509 |
+
}
|
nl_tasks/exp395/run_ex08/ft/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"T": 1.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"inference_mode": false,
|
| 6 |
+
"layers_to_transform": null,
|
| 7 |
+
"modules_to_save": null,
|
| 8 |
+
"num_rotations": 1,
|
| 9 |
+
"peft_type": "ROTATION",
|
| 10 |
+
"r": 16,
|
| 11 |
+
"revision": null,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"v_proj",
|
| 14 |
+
"q_proj"
|
| 15 |
+
],
|
| 16 |
+
"target_modules_to_skip": null,
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
nl_tasks/exp395/run_ex08/ft/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[PAD]": 32000
|
| 3 |
+
}
|
nl_tasks/exp395/run_ex08/ft/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"pad_token": {
|
| 17 |
+
"content": "[PAD]",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "</s>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
nl_tasks/exp395/run_ex08/ft/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
nl_tasks/exp395/run_ex08/ft/tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
nl_tasks/exp395/run_ex08/ft/tokenizer_config.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": null,
|
| 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 |
+
"32000": {
|
| 31 |
+
"content": "[PAD]",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false,
|
| 36 |
+
"special": true
|
| 37 |
+
}
|
| 38 |
+
},
|
| 39 |
+
"bos_token": "</s>",
|
| 40 |
+
"clean_up_tokenization_spaces": false,
|
| 41 |
+
"eos_token": "</s>",
|
| 42 |
+
"extra_special_tokens": {},
|
| 43 |
+
"legacy": false,
|
| 44 |
+
"model_max_length": 512,
|
| 45 |
+
"pad_token": "[PAD]",
|
| 46 |
+
"padding_side": "right",
|
| 47 |
+
"sp_model_kwargs": {},
|
| 48 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 49 |
+
"unk_token": "</s>",
|
| 50 |
+
"use_default_system_prompt": false
|
| 51 |
+
}
|
nl_tasks/exp395/run_ex08/ft2/adapter_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"T": 1.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"layers_to_transform": null,
|
| 7 |
+
"modules_to_save": null,
|
| 8 |
+
"num_rotations": 1,
|
| 9 |
+
"peft_type": "ROTATION",
|
| 10 |
+
"r": 16,
|
| 11 |
+
"revision": null,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"v_proj",
|
| 14 |
+
"q_proj"
|
| 15 |
+
],
|
| 16 |
+
"target_modules_to_skip": null,
|
| 17 |
+
"task_type": "CAUSAL_LM"
|
| 18 |
+
}
|
nl_tasks/exp395/run_ex08/ft2/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:343a959084dc5ab62addd82639a1a2f247ced9be6fcdfe7507654726d199f794
|
| 3 |
+
size 33602915
|
nl_tasks/exp395/run_ex08/trainer_state.json
ADDED
|
@@ -0,0 +1,509 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 9375,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.064,
|
| 14 |
+
"grad_norm": 0.3168371915817261,
|
| 15 |
+
"learning_rate": 9.99684221114305e-05,
|
| 16 |
+
"loss": 0.4089,
|
| 17 |
+
"step": 200
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.128,
|
| 21 |
+
"grad_norm": 0.36081254482269287,
|
| 22 |
+
"learning_rate": 9.973376564462874e-05,
|
| 23 |
+
"loss": 0.3014,
|
| 24 |
+
"step": 400
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.16,
|
| 28 |
+
"eval_loss": 0.2922876477241516,
|
| 29 |
+
"eval_runtime": 19.7057,
|
| 30 |
+
"eval_samples_per_second": 51.305,
|
| 31 |
+
"eval_steps_per_second": 0.812,
|
| 32 |
+
"step": 500
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.192,
|
| 36 |
+
"grad_norm": 0.3251103460788727,
|
| 37 |
+
"learning_rate": 9.927125570277145e-05,
|
| 38 |
+
"loss": 0.282,
|
| 39 |
+
"step": 600
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.256,
|
| 43 |
+
"grad_norm": 0.3563052713871002,
|
| 44 |
+
"learning_rate": 9.858301125867588e-05,
|
| 45 |
+
"loss": 0.2695,
|
| 46 |
+
"step": 800
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.32,
|
| 50 |
+
"grad_norm": 0.3129023015499115,
|
| 51 |
+
"learning_rate": 9.767218547973729e-05,
|
| 52 |
+
"loss": 0.2624,
|
| 53 |
+
"step": 1000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.32,
|
| 57 |
+
"eval_loss": 0.26842427253723145,
|
| 58 |
+
"eval_runtime": 19.4871,
|
| 59 |
+
"eval_samples_per_second": 51.881,
|
| 60 |
+
"eval_steps_per_second": 0.821,
|
| 61 |
+
"step": 1000
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.384,
|
| 65 |
+
"grad_norm": 0.34633564949035645,
|
| 66 |
+
"learning_rate": 9.654295128180494e-05,
|
| 67 |
+
"loss": 0.2603,
|
| 68 |
+
"step": 1200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.448,
|
| 72 |
+
"grad_norm": 0.32980990409851074,
|
| 73 |
+
"learning_rate": 9.520048221111679e-05,
|
| 74 |
+
"loss": 0.2544,
|
| 75 |
+
"step": 1400
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.48,
|
| 79 |
+
"eval_loss": 0.2564426064491272,
|
| 80 |
+
"eval_runtime": 19.5049,
|
| 81 |
+
"eval_samples_per_second": 51.833,
|
| 82 |
+
"eval_steps_per_second": 0.82,
|
| 83 |
+
"step": 1500
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.512,
|
| 87 |
+
"grad_norm": 0.32643768191337585,
|
| 88 |
+
"learning_rate": 9.365092874188177e-05,
|
| 89 |
+
"loss": 0.2489,
|
| 90 |
+
"step": 1600
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.576,
|
| 94 |
+
"grad_norm": 0.31819966435432434,
|
| 95 |
+
"learning_rate": 9.190139009810142e-05,
|
| 96 |
+
"loss": 0.2444,
|
| 97 |
+
"step": 1800
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.64,
|
| 101 |
+
"grad_norm": 0.3185544013977051,
|
| 102 |
+
"learning_rate": 8.995988172872798e-05,
|
| 103 |
+
"loss": 0.2429,
|
| 104 |
+
"step": 2000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.64,
|
| 108 |
+
"eval_loss": 0.2463328242301941,
|
| 109 |
+
"eval_runtime": 19.5247,
|
| 110 |
+
"eval_samples_per_second": 51.781,
|
| 111 |
+
"eval_steps_per_second": 0.819,
|
| 112 |
+
"step": 2000
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.704,
|
| 116 |
+
"grad_norm": 0.30173778533935547,
|
| 117 |
+
"learning_rate": 8.783529858517076e-05,
|
| 118 |
+
"loss": 0.2388,
|
| 119 |
+
"step": 2200
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.768,
|
| 123 |
+
"grad_norm": 0.2898229658603668,
|
| 124 |
+
"learning_rate": 8.553737436939324e-05,
|
| 125 |
+
"loss": 0.2366,
|
| 126 |
+
"step": 2400
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.8,
|
| 130 |
+
"eval_loss": 0.24057233333587646,
|
| 131 |
+
"eval_runtime": 19.4831,
|
| 132 |
+
"eval_samples_per_second": 51.891,
|
| 133 |
+
"eval_steps_per_second": 0.821,
|
| 134 |
+
"step": 2500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.832,
|
| 138 |
+
"grad_norm": 0.34637656807899475,
|
| 139 |
+
"learning_rate": 8.307663693930425e-05,
|
| 140 |
+
"loss": 0.234,
|
| 141 |
+
"step": 2600
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.896,
|
| 145 |
+
"grad_norm": 0.3339170217514038,
|
| 146 |
+
"learning_rate": 8.046436007575221e-05,
|
| 147 |
+
"loss": 0.2336,
|
| 148 |
+
"step": 2800
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.96,
|
| 152 |
+
"grad_norm": 0.2930004596710205,
|
| 153 |
+
"learning_rate": 7.771251183209993e-05,
|
| 154 |
+
"loss": 0.2265,
|
| 155 |
+
"step": 3000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.96,
|
| 159 |
+
"eval_loss": 0.23496957123279572,
|
| 160 |
+
"eval_runtime": 19.4688,
|
| 161 |
+
"eval_samples_per_second": 51.929,
|
| 162 |
+
"eval_steps_per_second": 0.822,
|
| 163 |
+
"step": 3000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.024,
|
| 167 |
+
"grad_norm": 0.3596031665802002,
|
| 168 |
+
"learning_rate": 7.483369970301454e-05,
|
| 169 |
+
"loss": 0.2246,
|
| 170 |
+
"step": 3200
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.088,
|
| 174 |
+
"grad_norm": 0.32522472739219666,
|
| 175 |
+
"learning_rate": 7.184111286368001e-05,
|
| 176 |
+
"loss": 0.2124,
|
| 177 |
+
"step": 3400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.12,
|
| 181 |
+
"eval_loss": 0.23045767843723297,
|
| 182 |
+
"eval_runtime": 19.488,
|
| 183 |
+
"eval_samples_per_second": 51.878,
|
| 184 |
+
"eval_steps_per_second": 0.821,
|
| 185 |
+
"step": 3500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.152,
|
| 189 |
+
"grad_norm": 0.331019788980484,
|
| 190 |
+
"learning_rate": 6.874846174406093e-05,
|
| 191 |
+
"loss": 0.2122,
|
| 192 |
+
"step": 3600
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.216,
|
| 196 |
+
"grad_norm": 0.3516997694969177,
|
| 197 |
+
"learning_rate": 6.556991521505634e-05,
|
| 198 |
+
"loss": 0.2112,
|
| 199 |
+
"step": 3800
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.28,
|
| 203 |
+
"grad_norm": 0.31302034854888916,
|
| 204 |
+
"learning_rate": 6.232003567432242e-05,
|
| 205 |
+
"loss": 0.2077,
|
| 206 |
+
"step": 4000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.28,
|
| 210 |
+
"eval_loss": 0.22687973082065582,
|
| 211 |
+
"eval_runtime": 19.5334,
|
| 212 |
+
"eval_samples_per_second": 51.758,
|
| 213 |
+
"eval_steps_per_second": 0.819,
|
| 214 |
+
"step": 4000
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.3439999999999999,
|
| 218 |
+
"grad_norm": 0.333069771528244,
|
| 219 |
+
"learning_rate": 5.901371232916676e-05,
|
| 220 |
+
"loss": 0.2102,
|
| 221 |
+
"step": 4200
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.408,
|
| 225 |
+
"grad_norm": 0.4020955562591553,
|
| 226 |
+
"learning_rate": 5.566609298217553e-05,
|
| 227 |
+
"loss": 0.208,
|
| 228 |
+
"step": 4400
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.44,
|
| 232 |
+
"eval_loss": 0.22374407947063446,
|
| 233 |
+
"eval_runtime": 19.772,
|
| 234 |
+
"eval_samples_per_second": 51.133,
|
| 235 |
+
"eval_steps_per_second": 0.809,
|
| 236 |
+
"step": 4500
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.472,
|
| 240 |
+
"grad_norm": 0.3483867645263672,
|
| 241 |
+
"learning_rate": 5.229251463209568e-05,
|
| 242 |
+
"loss": 0.2087,
|
| 243 |
+
"step": 4600
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.536,
|
| 247 |
+
"grad_norm": 0.3387017250061035,
|
| 248 |
+
"learning_rate": 4.8908433207921846e-05,
|
| 249 |
+
"loss": 0.2056,
|
| 250 |
+
"step": 4800
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 1.6,
|
| 254 |
+
"grad_norm": 0.3310563266277313,
|
| 255 |
+
"learning_rate": 4.5529352758108325e-05,
|
| 256 |
+
"loss": 0.2032,
|
| 257 |
+
"step": 5000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.6,
|
| 261 |
+
"eval_loss": 0.22034206986427307,
|
| 262 |
+
"eval_runtime": 19.5875,
|
| 263 |
+
"eval_samples_per_second": 51.615,
|
| 264 |
+
"eval_steps_per_second": 0.817,
|
| 265 |
+
"step": 5000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 1.6640000000000001,
|
| 269 |
+
"grad_norm": 0.3283143937587738,
|
| 270 |
+
"learning_rate": 4.2170754419323566e-05,
|
| 271 |
+
"loss": 0.2065,
|
| 272 |
+
"step": 5200
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.728,
|
| 276 |
+
"grad_norm": 0.32001566886901855,
|
| 277 |
+
"learning_rate": 3.884802549017425e-05,
|
| 278 |
+
"loss": 0.206,
|
| 279 |
+
"step": 5400
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.76,
|
| 283 |
+
"eval_loss": 0.21794956922531128,
|
| 284 |
+
"eval_runtime": 19.574,
|
| 285 |
+
"eval_samples_per_second": 51.65,
|
| 286 |
+
"eval_steps_per_second": 0.817,
|
| 287 |
+
"step": 5500
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.792,
|
| 291 |
+
"grad_norm": 0.3461301624774933,
|
| 292 |
+
"learning_rate": 3.557638893484501e-05,
|
| 293 |
+
"loss": 0.2039,
|
| 294 |
+
"step": 5600
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.8559999999999999,
|
| 298 |
+
"grad_norm": 0.358914315700531,
|
| 299 |
+
"learning_rate": 3.237083363963042e-05,
|
| 300 |
+
"loss": 0.2007,
|
| 301 |
+
"step": 5800
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.92,
|
| 305 |
+
"grad_norm": 0.3654290437698364,
|
| 306 |
+
"learning_rate": 2.9246045741886696e-05,
|
| 307 |
+
"loss": 0.2017,
|
| 308 |
+
"step": 6000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.92,
|
| 312 |
+
"eval_loss": 0.21639443933963776,
|
| 313 |
+
"eval_runtime": 19.5656,
|
| 314 |
+
"eval_samples_per_second": 51.672,
|
| 315 |
+
"eval_steps_per_second": 0.818,
|
| 316 |
+
"step": 6000
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.984,
|
| 320 |
+
"grad_norm": 0.3528446853160858,
|
| 321 |
+
"learning_rate": 2.6216341346016615e-05,
|
| 322 |
+
"loss": 0.2018,
|
| 323 |
+
"step": 6200
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 2.048,
|
| 327 |
+
"grad_norm": 0.36889269948005676,
|
| 328 |
+
"learning_rate": 2.3295600934747398e-05,
|
| 329 |
+
"loss": 0.1913,
|
| 330 |
+
"step": 6400
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 2.08,
|
| 334 |
+
"eval_loss": 0.21491438150405884,
|
| 335 |
+
"eval_runtime": 19.5386,
|
| 336 |
+
"eval_samples_per_second": 51.744,
|
| 337 |
+
"eval_steps_per_second": 0.819,
|
| 338 |
+
"step": 6500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.112,
|
| 342 |
+
"grad_norm": 0.35566315054893494,
|
| 343 |
+
"learning_rate": 2.049720577619374e-05,
|
| 344 |
+
"loss": 0.1873,
|
| 345 |
+
"step": 6600
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.176,
|
| 349 |
+
"grad_norm": 0.3847609758377075,
|
| 350 |
+
"learning_rate": 1.7833976618054676e-05,
|
| 351 |
+
"loss": 0.1876,
|
| 352 |
+
"step": 6800
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 2.24,
|
| 356 |
+
"grad_norm": 0.3727908730506897,
|
| 357 |
+
"learning_rate": 1.531811494981501e-05,
|
| 358 |
+
"loss": 0.185,
|
| 359 |
+
"step": 7000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 2.24,
|
| 363 |
+
"eval_loss": 0.21415850520133972,
|
| 364 |
+
"eval_runtime": 19.5534,
|
| 365 |
+
"eval_samples_per_second": 51.705,
|
| 366 |
+
"eval_steps_per_second": 0.818,
|
| 367 |
+
"step": 7000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 2.304,
|
| 371 |
+
"grad_norm": 0.3717604875564575,
|
| 372 |
+
"learning_rate": 1.296114710205592e-05,
|
| 373 |
+
"loss": 0.1874,
|
| 374 |
+
"step": 7200
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.368,
|
| 378 |
+
"grad_norm": 0.35382431745529175,
|
| 379 |
+
"learning_rate": 1.0773871438982197e-05,
|
| 380 |
+
"loss": 0.1906,
|
| 381 |
+
"step": 7400
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.4,
|
| 385 |
+
"eval_loss": 0.2131689041852951,
|
| 386 |
+
"eval_runtime": 19.5504,
|
| 387 |
+
"eval_samples_per_second": 51.712,
|
| 388 |
+
"eval_steps_per_second": 0.818,
|
| 389 |
+
"step": 7500
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.432,
|
| 393 |
+
"grad_norm": 0.39091038703918457,
|
| 394 |
+
"learning_rate": 8.766308886101405e-06,
|
| 395 |
+
"loss": 0.1895,
|
| 396 |
+
"step": 7600
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 2.496,
|
| 400 |
+
"grad_norm": 0.39373019337654114,
|
| 401 |
+
"learning_rate": 6.947657019710796e-06,
|
| 402 |
+
"loss": 0.1895,
|
| 403 |
+
"step": 7800
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 2.56,
|
| 407 |
+
"grad_norm": 0.3592659533023834,
|
| 408 |
+
"learning_rate": 5.326247928529187e-06,
|
| 409 |
+
"loss": 0.1878,
|
| 410 |
+
"step": 8000
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.56,
|
| 414 |
+
"eval_loss": 0.21238671243190765,
|
| 415 |
+
"eval_runtime": 19.5697,
|
| 416 |
+
"eval_samples_per_second": 51.662,
|
| 417 |
+
"eval_steps_per_second": 0.818,
|
| 418 |
+
"step": 8000
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 2.624,
|
| 422 |
+
"grad_norm": 0.36308494210243225,
|
| 423 |
+
"learning_rate": 3.90951004052949e-06,
|
| 424 |
+
"loss": 0.1859,
|
| 425 |
+
"step": 8200
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 2.6879999999999997,
|
| 429 |
+
"grad_norm": 0.3989940583705902,
|
| 430 |
+
"learning_rate": 2.703934089860627e-06,
|
| 431 |
+
"loss": 0.1878,
|
| 432 |
+
"step": 8400
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 2.7199999999999998,
|
| 436 |
+
"eval_loss": 0.2119983732700348,
|
| 437 |
+
"eval_runtime": 19.585,
|
| 438 |
+
"eval_samples_per_second": 51.621,
|
| 439 |
+
"eval_steps_per_second": 0.817,
|
| 440 |
+
"step": 8500
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.752,
|
| 444 |
+
"grad_norm": 0.3984169661998749,
|
| 445 |
+
"learning_rate": 1.7150433797803012e-06,
|
| 446 |
+
"loss": 0.1847,
|
| 447 |
+
"step": 8600
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 2.816,
|
| 451 |
+
"grad_norm": 0.4087405204772949,
|
| 452 |
+
"learning_rate": 9.473684778379677e-07,
|
| 453 |
+
"loss": 0.1881,
|
| 454 |
+
"step": 8800
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 2.88,
|
| 458 |
+
"grad_norm": 0.35457876324653625,
|
| 459 |
+
"learning_rate": 4.0442645924106404e-07,
|
| 460 |
+
"loss": 0.1872,
|
| 461 |
+
"step": 9000
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.88,
|
| 465 |
+
"eval_loss": 0.21190518140792847,
|
| 466 |
+
"eval_runtime": 19.5621,
|
| 467 |
+
"eval_samples_per_second": 51.682,
|
| 468 |
+
"eval_steps_per_second": 0.818,
|
| 469 |
+
"step": 9000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 2.944,
|
| 473 |
+
"grad_norm": 0.3945753574371338,
|
| 474 |
+
"learning_rate": 8.87047935002272e-08,
|
| 475 |
+
"loss": 0.1838,
|
| 476 |
+
"step": 9200
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 3.0,
|
| 480 |
+
"step": 9375,
|
| 481 |
+
"total_flos": 6.097062592512e+18,
|
| 482 |
+
"train_loss": 0.21853865112304688,
|
| 483 |
+
"train_runtime": 8852.7007,
|
| 484 |
+
"train_samples_per_second": 33.888,
|
| 485 |
+
"train_steps_per_second": 1.059
|
| 486 |
+
}
|
| 487 |
+
],
|
| 488 |
+
"logging_steps": 200,
|
| 489 |
+
"max_steps": 9375,
|
| 490 |
+
"num_input_tokens_seen": 0,
|
| 491 |
+
"num_train_epochs": 3,
|
| 492 |
+
"save_steps": 0,
|
| 493 |
+
"stateful_callbacks": {
|
| 494 |
+
"TrainerControl": {
|
| 495 |
+
"args": {
|
| 496 |
+
"should_epoch_stop": false,
|
| 497 |
+
"should_evaluate": false,
|
| 498 |
+
"should_log": false,
|
| 499 |
+
"should_save": true,
|
| 500 |
+
"should_training_stop": true
|
| 501 |
+
},
|
| 502 |
+
"attributes": {}
|
| 503 |
+
}
|
| 504 |
+
},
|
| 505 |
+
"total_flos": 6.097062592512e+18,
|
| 506 |
+
"train_batch_size": 32,
|
| 507 |
+
"trial_name": null,
|
| 508 |
+
"trial_params": null
|
| 509 |
+
}
|