Text Generation
Safetensors
Rust
RWKV
English
oicio-rs
ternary
matmul-free
cpu-only
1.58-bit
bitnet
bonsai
infinite-context
em-llm
reattention
recursive-agent-harness
rlm
rah
edge-ai
needle
hadamard
mlgru
mamba
liquid-neural-networks
turbovec
turboquant
t-mac
vec-lut
axon
consumer-hardware
better-quality
intelligence-density
Instructions to use deeprcurs/OICIO with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- RWKV
How to use deeprcurs/OICIO with RWKV:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
Upload models/BitNet-b1.58-2B-4T/config.json with huggingface_hub
Browse files
models/BitNet-b1.58-2B-4T/config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BitNetForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"auto_map": {
|
| 6 |
+
"AutoConfig": "configuration_bitnet.BitNetConfig",
|
| 7 |
+
"AutoModelForCausalLM": "modeling_bitnet.BitNetForCausalLM"
|
| 8 |
+
},
|
| 9 |
+
"bos_token_id": 128000,
|
| 10 |
+
"eos_token_id": 128001,
|
| 11 |
+
"hidden_act": "relu2",
|
| 12 |
+
"hidden_size": 2560,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 6912,
|
| 15 |
+
"max_position_embeddings": 4096,
|
| 16 |
+
"model_type": "bitnet",
|
| 17 |
+
"rms_norm_eps": 1e-05,
|
| 18 |
+
"num_attention_heads": 20,
|
| 19 |
+
"num_hidden_layers": 30,
|
| 20 |
+
"num_key_value_heads": 5,
|
| 21 |
+
"rope_theta": 500000.0,
|
| 22 |
+
"tie_word_embeddings": true,
|
| 23 |
+
"torch_dtype": "bfloat16",
|
| 24 |
+
"use_cache": true,
|
| 25 |
+
"vocab_size": 128256,
|
| 26 |
+
"quantization_config": {
|
| 27 |
+
"quant_method": "bitnet",
|
| 28 |
+
"linear_class": "autobitlinear",
|
| 29 |
+
"quantization_mode": "offline"
|
| 30 |
+
}
|
| 31 |
+
}
|