初始化项目,由ModelHub XC社区提供模型
Model: fpadovani/eng-latn-10mb-after-ppt-shuff-dyck-100mb-ckpt500_seed3407 Source: Original Platform
This commit is contained in:
1202
checkpoint-33000/added_tokens.json
Normal file
1202
checkpoint-33000/added_tokens.json
Normal file
File diff suppressed because it is too large
Load Diff
34
checkpoint-33000/config.json
Normal file
34
checkpoint-33000/config.json
Normal file
@@ -0,0 +1,34 @@
|
||||
{
|
||||
"activation_function": "gelu",
|
||||
"architectures": [
|
||||
"GPT2LMHeadModel"
|
||||
],
|
||||
"attn_pdrop": 0.1,
|
||||
"bos_token_id": 50000,
|
||||
"dtype": "bfloat16",
|
||||
"embd_pdrop": 0.1,
|
||||
"eos_token_id": 50001,
|
||||
"initializer_range": 0.02,
|
||||
"layer_norm_epsilon": 1e-05,
|
||||
"model_type": "gpt2",
|
||||
"n_ctx": 512,
|
||||
"n_embd": 512,
|
||||
"n_head": 8,
|
||||
"n_inner": 2048,
|
||||
"n_layer": 4,
|
||||
"n_positions": 512,
|
||||
"pad_token_id": 50002,
|
||||
"prefix": "[CLS]",
|
||||
"reorder_and_upcast_attn": false,
|
||||
"resid_pdrop": 0.1,
|
||||
"scale_attn_by_inverse_layer_idx": false,
|
||||
"scale_attn_weights": true,
|
||||
"summary_activation": null,
|
||||
"summary_first_dropout": 0.1,
|
||||
"summary_proj_to_labels": true,
|
||||
"summary_type": "cls_index",
|
||||
"summary_use_proj": true,
|
||||
"transformers_version": "4.56.2",
|
||||
"use_cache": true,
|
||||
"vocab_size": 51200
|
||||
}
|
||||
9
checkpoint-33000/generation_config.json
Normal file
9
checkpoint-33000/generation_config.json
Normal file
@@ -0,0 +1,9 @@
|
||||
{
|
||||
"_from_model_config": true,
|
||||
"bos_token_id": 50000,
|
||||
"eos_token_id": [
|
||||
50001
|
||||
],
|
||||
"pad_token_id": 50002,
|
||||
"transformers_version": "4.56.2"
|
||||
}
|
||||
3
checkpoint-33000/model.safetensors
Normal file
3
checkpoint-33000/model.safetensors
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:18b1678465bbce2e38853cfe465a1fa3b84b5395ee700a28d1fb730c29782c31
|
||||
size 78179408
|
||||
3
checkpoint-33000/rng_state.pth
Normal file
3
checkpoint-33000/rng_state.pth
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:08c3e42c5ad5111c4f2ad389d83a3d1d34fa708050a4bc5d0070cbcb3bc842b0
|
||||
size 14645
|
||||
3
checkpoint-33000/scheduler.pt
Normal file
3
checkpoint-33000/scheduler.pt
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:1b73e81f861200f0fb37b5a16b35a2d40e0e9dc76df1147e576922c8956121df
|
||||
size 1465
|
||||
1249
checkpoint-33000/special_tokens_map.json
Normal file
1249
checkpoint-33000/special_tokens_map.json
Normal file
File diff suppressed because it is too large
Load Diff
3
checkpoint-33000/spiece.model
Normal file
3
checkpoint-33000/spiece.model
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:5ca5e38fbedf0b9d03dfb93cd1435efd6bffda4d8378011a7c97ec1214d7d51e
|
||||
size 1128248
|
||||
10825
checkpoint-33000/tokenizer_config.json
Normal file
10825
checkpoint-33000/tokenizer_config.json
Normal file
File diff suppressed because it is too large
Load Diff
66155
checkpoint-33000/trainer_state.json
Normal file
66155
checkpoint-33000/trainer_state.json
Normal file
File diff suppressed because it is too large
Load Diff
3
checkpoint-33000/training_args.bin
Normal file
3
checkpoint-33000/training_args.bin
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:5261e2ea905b60b5b7b79423215dec9e01e05e9719e49b44602956e99ceec063
|
||||
size 6353
|
||||
Reference in New Issue
Block a user