Update tiny models for CTRLLMHeadModel
Browse files- config.json +24 -0
- generation_config.json +8 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- tokenizer_config.json +18 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"CTRLLMHeadModel"
|
| 4 |
+
],
|
| 5 |
+
"bos_token_id": null,
|
| 6 |
+
"dff": 37,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"embd_pdrop": 0.1,
|
| 9 |
+
"eos_token_id": null,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"is_decoder": true,
|
| 12 |
+
"layer_norm_epsilon": 1e-06,
|
| 13 |
+
"model_type": "ctrl",
|
| 14 |
+
"n_embd": 32,
|
| 15 |
+
"n_head": 4,
|
| 16 |
+
"n_layer": 2,
|
| 17 |
+
"n_positions": 512,
|
| 18 |
+
"pad_token_id": 246533,
|
| 19 |
+
"resid_pdrop": 0.1,
|
| 20 |
+
"tie_word_embeddings": true,
|
| 21 |
+
"transformers_version": "5.16.0.dev0",
|
| 22 |
+
"use_cache": true,
|
| 23 |
+
"vocab_size": 246534
|
| 24 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"output_attentions": false,
|
| 4 |
+
"output_hidden_states": false,
|
| 5 |
+
"pad_token_id": 246533,
|
| 6 |
+
"transformers_version": "5.16.0.dev0",
|
| 7 |
+
"use_cache": true
|
| 8 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2efc6d00568cdb26e0cd07b5ea2a80831f67b92ff624912978ac1f3d45ac032
|
| 3 |
+
size 32600752
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"246532": {
|
| 4 |
+
"content": "<unk>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
}
|
| 11 |
+
},
|
| 12 |
+
"backend": "custom",
|
| 13 |
+
"is_local": false,
|
| 14 |
+
"local_files_only": false,
|
| 15 |
+
"model_max_length": 512,
|
| 16 |
+
"tokenizer_class": "CTRLTokenizer",
|
| 17 |
+
"unk_token": "<unk>"
|
| 18 |
+
}
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|