初始化项目,由ModelHub XC社区提供模型
Model: fpadovani/eng-latn-10mb-10mb_seed3407 Source: Original Platform
This commit is contained in:
1202
checkpoint-2500/added_tokens.json
Normal file
1202
checkpoint-2500/added_tokens.json
Normal file
File diff suppressed because it is too large
Load Diff
34
checkpoint-2500/config.json
Normal file
34
checkpoint-2500/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-2500/generation_config.json
Normal file
9
checkpoint-2500/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-2500/model.safetensors
Normal file
3
checkpoint-2500/model.safetensors
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:2165e15ffff5719ac6c5cc686361ab3bb9304e8afa8b8b99ebc63b0582c4e587
|
||||
size 78179408
|
||||
3
checkpoint-2500/rng_state.pth
Normal file
3
checkpoint-2500/rng_state.pth
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:a8598ff184c9aec8b2dddac53b7ee46383aa933b85b3869a920c69fef4b8f1c5
|
||||
size 14645
|
||||
3
checkpoint-2500/scheduler.pt
Normal file
3
checkpoint-2500/scheduler.pt
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:8f148cc9a47d9c16b1ad183d80a72602e22cab065766353fcda37d3e0628671e
|
||||
size 1465
|
||||
1249
checkpoint-2500/special_tokens_map.json
Normal file
1249
checkpoint-2500/special_tokens_map.json
Normal file
File diff suppressed because it is too large
Load Diff
3
checkpoint-2500/spiece.model
Normal file
3
checkpoint-2500/spiece.model
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:f0cc7d2d7e6989f700f65432cac2fcb2a9e872d11d2c9166338a1cc423019e6e
|
||||
size 1100651
|
||||
10825
checkpoint-2500/tokenizer_config.json
Normal file
10825
checkpoint-2500/tokenizer_config.json
Normal file
File diff suppressed because it is too large
Load Diff
5089
checkpoint-2500/trainer_state.json
Normal file
5089
checkpoint-2500/trainer_state.json
Normal file
File diff suppressed because it is too large
Load Diff
3
checkpoint-2500/training_args.bin
Normal file
3
checkpoint-2500/training_args.bin
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:27ec90e769d02d3d09ae90eeffefc154f388df2af968c971066a38b52f7bf716
|
||||
size 6289
|
||||
Reference in New Issue
Block a user