update model card README.md
Browse files
README.md
CHANGED
@@ -36,10 +36,10 @@ More information needed
|
|
36 |
|
37 |
The following hyperparameters were used during training:
|
38 |
- learning_rate: 0.0005
|
39 |
-
- train_batch_size:
|
40 |
-
- eval_batch_size:
|
41 |
- seed: 42
|
42 |
-
- gradient_accumulation_steps:
|
43 |
- total_train_batch_size: 128
|
44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
45 |
- lr_scheduler_type: linear
|
@@ -60,7 +60,7 @@ The following hyperparameters were used during training:
|
|
60 |
'is_split_by_sentences': True,
|
61 |
'skip_tokens': 1649999872},
|
62 |
'generation': {'batch_size': 128,
|
63 |
-
'every_n_steps':
|
64 |
'force_call_on': [12588],
|
65 |
'metrics_configs': [{}, {'n': 1}, {}],
|
66 |
'scenario_configs': [{'display_as_html': True,
|
@@ -88,7 +88,7 @@ The following hyperparameters were used during training:
|
|
88 |
'prompts_path': 'resources/functions_csnet.jsonl',
|
89 |
'use_prompt_for_scoring': True}],
|
90 |
'scorer_config': {}},
|
91 |
-
'kl_gpt3_callback': {'every_n_steps':
|
92 |
'force_call_on': [12588],
|
93 |
'gpt3_kwargs': {'model_name': 'code-cushman-001'},
|
94 |
'max_tokens': 64,
|
@@ -96,7 +96,8 @@ The following hyperparameters were used during training:
|
|
96 |
'model': {'from_scratch': False,
|
97 |
'gpt2_config_kwargs': {'reorder_and_upcast_attn': True,
|
98 |
'scale_attn_by': True},
|
99 |
-
'model_kwargs': {'revision': 'c38e2b6acf17781918d39a310ee1adc4674a8225'
|
|
|
100 |
'path_or_name': 'kejian/mighty-rwr'},
|
101 |
'objective': {'alpha': 1, 'beta': 10, 'name': 'AWR'},
|
102 |
'tokenizer': {'path_or_name': 'codeparrot/codeparrot-small'},
|
@@ -122,4 +123,4 @@ The following hyperparameters were used during training:
|
|
122 |
'weight_decay': 0.1}}
|
123 |
|
124 |
# Wandb URL:
|
125 |
-
https://wandb.ai/kejian/uncategorized/runs/
|
36 |
|
37 |
The following hyperparameters were used during training:
|
38 |
- learning_rate: 0.0005
|
39 |
+
- train_batch_size: 64
|
40 |
+
- eval_batch_size: 32
|
41 |
- seed: 42
|
42 |
+
- gradient_accumulation_steps: 2
|
43 |
- total_train_batch_size: 128
|
44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
45 |
- lr_scheduler_type: linear
|
60 |
'is_split_by_sentences': True,
|
61 |
'skip_tokens': 1649999872},
|
62 |
'generation': {'batch_size': 128,
|
63 |
+
'every_n_steps': 512,
|
64 |
'force_call_on': [12588],
|
65 |
'metrics_configs': [{}, {'n': 1}, {}],
|
66 |
'scenario_configs': [{'display_as_html': True,
|
88 |
'prompts_path': 'resources/functions_csnet.jsonl',
|
89 |
'use_prompt_for_scoring': True}],
|
90 |
'scorer_config': {}},
|
91 |
+
'kl_gpt3_callback': {'every_n_steps': 512,
|
92 |
'force_call_on': [12588],
|
93 |
'gpt3_kwargs': {'model_name': 'code-cushman-001'},
|
94 |
'max_tokens': 64,
|
96 |
'model': {'from_scratch': False,
|
97 |
'gpt2_config_kwargs': {'reorder_and_upcast_attn': True,
|
98 |
'scale_attn_by': True},
|
99 |
+
'model_kwargs': {'revision': 'c38e2b6acf17781918d39a310ee1adc4674a8225',
|
100 |
+
'value_head_config': {'is_detached': False}},
|
101 |
'path_or_name': 'kejian/mighty-rwr'},
|
102 |
'objective': {'alpha': 1, 'beta': 10, 'name': 'AWR'},
|
103 |
'tokenizer': {'path_or_name': 'codeparrot/codeparrot-small'},
|
123 |
'weight_decay': 0.1}}
|
124 |
|
125 |
# Wandb URL:
|
126 |
+
https://wandb.ai/kejian/uncategorized/runs/39mf4btg
|