Instructions to use beast33/92dd539b-2218-4565-9bb6-ce7dc3b376a1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use beast33/92dd539b-2218-4565-9bb6-ce7dc3b376a1 with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("zake7749/gemma-2-2b-it-chinese-kyara-dpo") model = PeftModel.from_pretrained(base_model, "beast33/92dd539b-2218-4565-9bb6-ce7dc3b376a1") - Notebooks
- Google Colab
- Kaggle
Download config.json from beast33/92dd539b-2218-4565-9bb6-ce7dc3b376a1: direct link, hf CLI and curl.
- Browser
- Download file 922 Bytes
-
https://huggingface.co/beast33/92dd539b-2218-4565-9bb6-ce7dc3b376a1/resolve/927a1b12a8b7ca5c6869c3ee6441c5ce6dce0b86/config.json
- Command line
-
hf download hf://beast33/92dd539b-2218-4565-9bb6-ce7dc3b376a1@927a1b12a8b7ca5c6869c3ee6441c5ce6dce0b86/config.json
-
curl -L -o config.json https://huggingface.co/beast33/92dd539b-2218-4565-9bb6-ce7dc3b376a1/resolve/927a1b12a8b7ca5c6869c3ee6441c5ce6dce0b86/config.json
922 Bytes
| { | |
| "_attn_implementation_autoset": true, | |
| "_name_or_path": "zake7749/gemma-2-2b-it-chinese-kyara-dpo", | |
| "architectures": [ | |
| "Gemma2ForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "attn_logit_softcapping": 50.0, | |
| "bos_token_id": 2, | |
| "cache_implementation": "hybrid", | |
| "eos_token_id": 1, | |
| "final_logit_softcapping": 30.0, | |
| "head_dim": 256, | |
| "hidden_act": "gelu_pytorch_tanh", | |
| "hidden_activation": "gelu_pytorch_tanh", | |
| "hidden_size": 2304, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 9216, | |
| "max_position_embeddings": 8192, | |
| "model_type": "gemma2", | |
| "num_attention_heads": 8, | |
| "num_hidden_layers": 26, | |
| "num_key_value_heads": 4, | |
| "pad_token_id": 0, | |
| "query_pre_attn_scalar": 256, | |
| "rms_norm_eps": 1e-06, | |
| "rope_theta": 10000.0, | |
| "sliding_window": 4096, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.46.0", | |
| "use_cache": false, | |
| "vocab_size": 256000 | |
| } | |