初始化项目,由ModelHub XC社区提供模型
Model: geasslabs/CC-Zeta-0 Source: Original Platform
This commit is contained in:
31
checkpoint-6670/config.json
Normal file
31
checkpoint-6670/config.json
Normal file
@@ -0,0 +1,31 @@
|
||||
{
|
||||
"architectures": [
|
||||
"MistralForCausalLM"
|
||||
],
|
||||
"attention_dropout": 0.0,
|
||||
"bos_token_id": 1,
|
||||
"dtype": "bfloat16",
|
||||
"eos_token_id": 2,
|
||||
"head_dim": 128,
|
||||
"hidden_act": "silu",
|
||||
"hidden_size": 4096,
|
||||
"initializer_range": 0.02,
|
||||
"intermediate_size": 14336,
|
||||
"max_position_embeddings": 32768,
|
||||
"model_type": "mistral",
|
||||
"num_attention_heads": 32,
|
||||
"num_hidden_layers": 32,
|
||||
"num_key_value_heads": 8,
|
||||
"pad_token_id": 770,
|
||||
"rms_norm_eps": 1e-05,
|
||||
"rope_parameters": {
|
||||
"rope_theta": 1000000.0,
|
||||
"rope_type": "default"
|
||||
},
|
||||
"sliding_window": null,
|
||||
"tie_word_embeddings": false,
|
||||
"transformers_version": "5.2.0",
|
||||
"unsloth_version": "2024.9",
|
||||
"use_cache": false,
|
||||
"vocab_size": 32768
|
||||
}
|
||||
8
checkpoint-6670/generation_config.json
Normal file
8
checkpoint-6670/generation_config.json
Normal file
@@ -0,0 +1,8 @@
|
||||
{
|
||||
"_from_model_config": true,
|
||||
"bos_token_id": 1,
|
||||
"eos_token_id": 2,
|
||||
"max_length": 32768,
|
||||
"pad_token_id": 770,
|
||||
"transformers_version": "5.2.0"
|
||||
}
|
||||
3
checkpoint-6670/model.safetensors
Normal file
3
checkpoint-6670/model.safetensors
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:c783ca6294e81310e3871e2e03b78175ea4fc801fbc87503589c8604ef906dbd
|
||||
size 14496081136
|
||||
3
checkpoint-6670/optimizer.pt
Normal file
3
checkpoint-6670/optimizer.pt
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:e9404cc9fb1d142d2bdce6da1d557955781e7e5d4d3b02749943532b73da3096
|
||||
size 15524017779
|
||||
3
checkpoint-6670/rng_state.pth
Normal file
3
checkpoint-6670/rng_state.pth
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:61c19bab1174704a4a4441475683bf1270277af15d2e2c95e964789128e482c4
|
||||
size 14645
|
||||
3
checkpoint-6670/scheduler.pt
Normal file
3
checkpoint-6670/scheduler.pt
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:5138e533ba751b47b6334d8a30f897acd77cc8655701ba54955421a238ea8c89
|
||||
size 1465
|
||||
275738
checkpoint-6670/tokenizer.json
Normal file
275738
checkpoint-6670/tokenizer.json
Normal file
File diff suppressed because it is too large
Load Diff
17
checkpoint-6670/tokenizer_config.json
Normal file
17
checkpoint-6670/tokenizer_config.json
Normal file
@@ -0,0 +1,17 @@
|
||||
{
|
||||
"add_prefix_space": true,
|
||||
"backend": "tokenizers",
|
||||
"bos_token": "<s>",
|
||||
"clean_up_tokenization_spaces": false,
|
||||
"eos_token": "</s>",
|
||||
"is_local": false,
|
||||
"legacy": false,
|
||||
"model_max_length": 1000000000000000019884624838656,
|
||||
"pad_token": "</s>",
|
||||
"padding_side": "right",
|
||||
"sp_model_kwargs": {},
|
||||
"spaces_between_special_tokens": false,
|
||||
"tokenizer_class": "TokenizersBackend",
|
||||
"unk_token": "<unk>",
|
||||
"use_default_system_prompt": false
|
||||
}
|
||||
4703
checkpoint-6670/trainer_state.json
Normal file
4703
checkpoint-6670/trainer_state.json
Normal file
File diff suppressed because it is too large
Load Diff
3
checkpoint-6670/training_args.bin
Normal file
3
checkpoint-6670/training_args.bin
Normal file
@@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:8f2579f1f92493888ad963448c2ba4b2d2e3c9ccbdbf0bc5743da86c84a49f91
|
||||
size 5201
|
||||
Reference in New Issue
Block a user