初始化项目,由ModelHub XC社区提供模型
Model: fpadovani/eus-latn-10mb-after-ppt-shuff-dyck-100mb-ckpt500_seed10 Source: Original Platform
This commit is contained in:
1202
checkpoint-21000/added_tokens.json
Normal file
1202
checkpoint-21000/added_tokens.json
Normal file
File diff suppressed because it is too large
Load Diff
34
checkpoint-21000/config.json
Normal file
34
checkpoint-21000/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-21000/generation_config.json
Normal file
9
checkpoint-21000/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-21000/model.safetensors
Normal file
3
checkpoint-21000/model.safetensors
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:905a3b0a10fba4308b1a191cc25d40a114fb766914ba80e549f040bbf18968c7
|
||||
size 78179408
|
||||
3
checkpoint-21000/rng_state.pth
Normal file
3
checkpoint-21000/rng_state.pth
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:2f5e42525f87b330b6a2f1bccea09fc991309116623486135341eb1faf3f0bcb
|
||||
size 14645
|
||||
1249
checkpoint-21000/special_tokens_map.json
Normal file
1249
checkpoint-21000/special_tokens_map.json
Normal file
File diff suppressed because it is too large
Load Diff
3
checkpoint-21000/spiece.model
Normal file
3
checkpoint-21000/spiece.model
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:3f141e2be9b9a4b8542400c72efca06b1a760fba52788ed65a0fafdf9382f967
|
||||
size 1172370
|
||||
10825
checkpoint-21000/tokenizer_config.json
Normal file
10825
checkpoint-21000/tokenizer_config.json
Normal file
File diff suppressed because it is too large
Load Diff
42111
checkpoint-21000/trainer_state.json
Normal file
42111
checkpoint-21000/trainer_state.json
Normal file
File diff suppressed because it is too large
Load Diff
3
checkpoint-21000/training_args.bin
Normal file
3
checkpoint-21000/training_args.bin
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:dcc749efb0f2659971258dde029719f80fcedad43ca4c8be9bb4e926af03fa8a
|
||||
size 6353
|
||||
Reference in New Issue
Block a user