[LLM-paddle] add llama1-7b pretrain with callback #239
update config
- Loading branch information
commit af2fd2f83fd560b8a326dc6b8c445eb4e814efad
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1 +1 @@ | ||
| ../llama1_7B/paddle/ |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Empty file.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,43 @@ | ||
| # model info | ||
| model_name_or_path: str = "facebook/llama-13b" | ||
| tokenizer_name_or_path: str = "facebook/llama-13b" | ||
| continue_training = 0 | ||
| split = "998,1,1" | ||
| max_seq_length = 2048 | ||
|
|
||
| # training info | ||
| dataloader_num_workers = 1 | ||
| max_steps = 512 | ||
| save_steps = 10000 | ||
| eval_steps = 10000 | ||
| learning_rate = 3e-4 | ||
| min_learning_rate = 3e-5 | ||
| warmup_steps = 2000 | ||
| weight_decay = 0.1 | ||
| lr_scheduler_type = "cosine" | ||
| adam_beta1 = 0.9 | ||
| adam_beta2 = 0.95 | ||
| adam_epsilon = 1e-06 | ||
| max_grad_norm = 1.0 | ||
| target_loss = 1.0 | ||
| target_ppl = 0.6 | ||
| logging_steps = 1 | ||
| log_freq = 1 | ||
| seed = 42 | ||
|
|
||
| # for parallel | ||
| per_device_train_batch_size = 4 | ||
| per_device_eval_batch_size = 1 | ||
| tensor_parallel_degree = 1 | ||
| pipeline_parallel_degree = 1 | ||
| sharding_parallel_degree = 8 | ||
| gradient_accumulation_steps = 32 | ||
| use_flash_attention = 1 | ||
| fuse_attention_qkv = 0 | ||
| use_fused_rms_norm = 1 | ||
| fp16 = True | ||
| fp16_opt_level = "O2" | ||
| scale_loss = 32768 | ||
| sharding = "stage2" | ||
| recompute = False | ||
| recompute_granularity = "full" |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,43 @@ | ||
| # model info | ||
| model_name_or_path: str = "facebook/llama-13b" | ||
| tokenizer_name_or_path: str = "facebook/llama-13b" | ||
| continue_training = 0 | ||
| split = "998,1,1" | ||
| max_seq_length = 2048 | ||
|
|
||
| # training info | ||
| dataloader_num_workers = 1 | ||
| max_steps = 512 | ||
| save_steps = 10000 | ||
| eval_steps = 10000 | ||
| learning_rate = 3e-4 | ||
| min_learning_rate = 3e-5 | ||
| warmup_steps = 2000 | ||
| weight_decay = 0.1 | ||
| lr_scheduler_type = "cosine" | ||
| adam_beta1 = 0.9 | ||
| adam_beta2 = 0.95 | ||
| adam_epsilon = 1e-06 | ||
| max_grad_norm = 1.0 | ||
| target_loss = 1.0 | ||
| target_ppl = 0.6 | ||
| logging_steps = 1 | ||
| log_freq = 1 | ||
| seed = 42 | ||
|
|
||
| # for parallel | ||
| per_device_train_batch_size = 4 | ||
| per_device_eval_batch_size = 1 | ||
| tensor_parallel_degree = 2 | ||
| pipeline_parallel_degree = 1 | ||
| sharding_parallel_degree = 4 | ||
| gradient_accumulation_steps = 64 | ||
| use_flash_attention = 1 | ||
| fuse_attention_qkv = 0 | ||
| use_fused_rms_norm = 1 | ||
| fp16 = True | ||
| fp16_opt_level = "O2" | ||
| scale_loss = 32768 | ||
| sharding = "stage1" | ||
| recompute = False | ||
| recompute_granularity = "full" |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,43 @@ | ||
| # model info | ||
| model_name_or_path: str = "facebook/llama-13b" | ||
| tokenizer_name_or_path: str = "facebook/llama-13b" | ||
| continue_training = 0 | ||
| split = "998,1,1" | ||
| max_seq_length = 2048 | ||
|
|
||
| # training info | ||
| dataloader_num_workers = 1 | ||
| max_steps = 512 | ||
| save_steps = 10000 | ||
| eval_steps = 10000 | ||
| learning_rate = 3e-4 | ||
| min_learning_rate = 3e-5 | ||
| warmup_steps = 2000 | ||
| weight_decay = 0.1 | ||
| lr_scheduler_type = "cosine" | ||
| adam_beta1 = 0.9 | ||
| adam_beta2 = 0.95 | ||
| adam_epsilon = 1e-06 | ||
| max_grad_norm = 1.0 | ||
| target_loss = 1.0 | ||
| target_ppl = 0.6 | ||
| logging_steps = 1 | ||
| log_freq = 1 | ||
| seed = 42 | ||
|
|
||
| # for parallel | ||
| per_device_train_batch_size = 4 | ||
| per_device_eval_batch_size = 1 | ||
| tensor_parallel_degree = 2 | ||
| pipeline_parallel_degree = 1 | ||
| sharding_parallel_degree = 4 | ||
| gradient_accumulation_steps = 64 | ||
| use_flash_attention = 1 | ||
| fuse_attention_qkv = 0 | ||
| use_fused_rms_norm = 1 | ||
| fp16 = True | ||
| fp16_opt_level = "O2" | ||
| scale_loss = 32768 | ||
| sharding = "stage2" | ||
| recompute = False | ||
| recompute_granularity = "full" |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,43 @@ | ||
| # model info | ||
| model_name_or_path: str = "facebook/llama-13b" | ||
| tokenizer_name_or_path: str = "facebook/llama-13b" | ||
| continue_training = 0 | ||
| split = "998,1,1" | ||
| max_seq_length = 2048 | ||
|
|
||
| # training info | ||
| dataloader_num_workers = 1 | ||
| max_steps = 512 | ||
| save_steps = 10000 | ||
| eval_steps = 10000 | ||
| learning_rate = 3e-4 | ||
| min_learning_rate = 3e-5 | ||
| warmup_steps = 2000 | ||
| weight_decay = 0.1 | ||
| lr_scheduler_type = "cosine" | ||
| adam_beta1 = 0.9 | ||
| adam_beta2 = 0.95 | ||
| adam_epsilon = 1e-06 | ||
| max_grad_norm = 1.0 | ||
| target_loss = 1.0 | ||
| target_ppl = 0.6 | ||
| logging_steps = 1 | ||
| log_freq = 1 | ||
| seed = 42 | ||
|
|
||
| # for parallel | ||
| per_device_train_batch_size = 4 | ||
| per_device_eval_batch_size = 1 | ||
| tensor_parallel_degree = 2 | ||
| pipeline_parallel_degree = 4 | ||
| sharding_parallel_degree = 1 | ||
| gradient_accumulation_steps = 256 | ||
| use_flash_attention = 1 | ||
| fuse_attention_qkv = 0 | ||
| use_fused_rms_norm = 1 | ||
| fp16 = True | ||
| fp16_opt_level = "O2" | ||
| scale_loss = 32768 | ||
| sharding = "stage1" | ||
| recompute = False | ||
| recompute_granularity = "full" |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1 +1 @@ | ||
| ../../llama1_7B-paddle/config/environment_variables.sh |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1 +1 @@ | ||
| ../../llama1_7B-paddle/config/requirements.txt |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,43 @@ | ||
| # model info | ||
| model_name_or_path: str = "facebook/llama-7b" | ||
| tokenizer_name_or_path: str = "facebook/llama-7b" | ||
| continue_training = 0 | ||
| split = "998,1,1" | ||
| max_seq_length = 2048 | ||
|
|
||
| # training info | ||
| dataloader_num_workers = 1 | ||
| max_steps = 512 | ||
| save_steps = 10000 | ||
| eval_steps = 10000 | ||
| learning_rate = 3e-4 | ||
| min_learning_rate = 3e-5 | ||
| warmup_steps = 2000 | ||
| weight_decay = 0.1 | ||
| lr_scheduler_type = "cosine" | ||
| adam_beta1 = 0.9 | ||
| adam_beta2 = 0.95 | ||
| adam_epsilon = 1e-06 | ||
| max_grad_norm = 1.0 | ||
| target_loss = 1.0 | ||
| target_ppl = 0.6 | ||
| logging_steps = 1 | ||
| log_freq = 1 | ||
| seed = 42 | ||
|
|
||
| # for parallel | ||
| per_device_train_batch_size = 4 | ||
| per_device_eval_batch_size = 1 | ||
| tensor_parallel_degree = 1 | ||
| pipeline_parallel_degree = 1 | ||
| sharding_parallel_degree = 8 | ||
| gradient_accumulation_steps = 32 | ||
| use_flash_attention = 1 | ||
| fuse_attention_qkv = 0 | ||
| use_fused_rms_norm = 1 | ||
| fp16 = True | ||
| fp16_opt_level = "O2" | ||
| scale_loss = 32768 | ||
| sharding = "stage2" | ||
| recompute = False | ||
| recompute_granularity = "full" |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,43 @@ | ||
| # model info | ||
| model_name_or_path: str = "facebook/llama-7b" | ||
| tokenizer_name_or_path: str = "facebook/llama-7b" | ||
| continue_training = 0 | ||
| split = "998,1,1" | ||
| max_seq_length = 2048 | ||
|
|
||
| # training info | ||
| dataloader_num_workers = 1 | ||
| max_steps = 512 | ||
| save_steps = 10000 | ||
| eval_steps = 10000 | ||
| learning_rate = 3e-4 | ||
| min_learning_rate = 3e-5 | ||
| warmup_steps = 2000 | ||
| weight_decay = 0.1 | ||
| lr_scheduler_type = "cosine" | ||
| adam_beta1 = 0.9 | ||
| adam_beta2 = 0.95 | ||
| adam_epsilon = 1e-06 | ||
| max_grad_norm = 1.0 | ||
| target_loss = 1.0 | ||
| target_ppl = 0.6 | ||
| logging_steps = 1 | ||
| log_freq = 1 | ||
| seed = 42 | ||
|
|
||
| # for parallel | ||
| per_device_train_batch_size = 4 | ||
| per_device_eval_batch_size = 1 | ||
| tensor_parallel_degree = 2 | ||
| pipeline_parallel_degree = 1 | ||
| sharding_parallel_degree = 4 | ||
| gradient_accumulation_steps = 64 | ||
| use_flash_attention = 1 | ||
| fuse_attention_qkv = 0 | ||
| use_fused_rms_norm = 1 | ||
| fp16 = True | ||
| fp16_opt_level = "O2" | ||
| scale_loss = 32768 | ||
| sharding = "stage1" | ||
| recompute = False | ||
| recompute_granularity = "full" |
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
请仿照其他case的格式书写一下llama1_7B预训练任务，包括模型描述、数据集下载、数据集处理脚本、代码来源开源协议等