Instructions to use kaanino/gpt-mha-RoPE with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use kaanino/gpt-mha-RoPE with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import GPT model = GPT.from_pretrained("kaanino/gpt-mha-RoPE", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Training in progress, step 107
Browse files- config.json +18 -0
- model.safetensors +3 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"GPT"
|
| 4 |
+
],
|
| 5 |
+
"dropout": 0.1,
|
| 6 |
+
"dtype": "float32",
|
| 7 |
+
"max_seq_length": 128,
|
| 8 |
+
"model_type": "gpt",
|
| 9 |
+
"n_embd": 64,
|
| 10 |
+
"n_head": 4,
|
| 11 |
+
"n_kv_head": 4,
|
| 12 |
+
"n_layer": 6,
|
| 13 |
+
"norm_type": "rmsnorm",
|
| 14 |
+
"pos_enc_type": "relative",
|
| 15 |
+
"transformers_version": "5.5.4",
|
| 16 |
+
"use_cache": false,
|
| 17 |
+
"vocab_size": 50257
|
| 18 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:662a7fb97cffa51f9db7e2ed75993628999c13fcd6118a8b40d879acf0a7cce1
|
| 3 |
+
size 27422560
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:522c81faacddcc3b8fc690db058c00e834ba96257d1b7a13ce5537e59c6f4084
|
| 3 |
+
size 5201
|