Feature Extraction
Transformers
Safetensors
microglot
biology
genomics
dna
microbiology
mixture-of-experts
species-conditioning
custom_code
Instructions to use athanzli/MicroGlot with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use athanzli/MicroGlot with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="athanzli/MicroGlot", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("athanzli/MicroGlot", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download config.json from athanzli/MicroGlot: direct link, hf CLI and curl.
- Browser
- Download file 1.9 kB
-
https://huggingface.co/athanzli/MicroGlot/resolve/main/config.json
- Command line
-
hf download hf://athanzli/MicroGlot/config.json
-
curl -L -o config.json https://huggingface.co/athanzli/MicroGlot/resolve/main/config.json
1.9 kB
| { | |
| "architectures": [ | |
| "MicroGlotModel" | |
| ], | |
| "attention_dropout_prob": 0.0, | |
| "attn_output_dropout_prob": 0.0, | |
| "auto_map": { | |
| "AutoConfig": "modeling_microglot.MicroGlotConfig", | |
| "AutoModel": "modeling_microglot.MicroGlotModel", | |
| "AutoModelForCausalLM": "modeling_microglot.MicroGlotForCausalLM" | |
| }, | |
| "bos_token_id": 1, | |
| "classifier_dropout_prob": 0.0, | |
| "eos_token_id": 2, | |
| "ffn_dropout_prob": 0.0, | |
| "gate_softmax_over_all_experts": true, | |
| "gradient_checkpointing": false, | |
| "hidden_size": 1024, | |
| "init_scale": 0.1, | |
| "intermediate_size": 2816, | |
| "max_trained_length": 8192, | |
| "model_type": "microglot", | |
| "moe_capacity_factor": 1.25, | |
| "moe_dispatch": "sorted", | |
| "moe_drop_tokens": true, | |
| "moe_eval_capacity_factor": 2.0, | |
| "moe_expert_dropout_prob": 0.0, | |
| "moe_layer_experts": [ | |
| 4, | |
| 64, | |
| 4, | |
| 32, | |
| 4, | |
| 16, | |
| 4, | |
| 8, | |
| 4, | |
| 8, | |
| 4, | |
| 8, | |
| 4, | |
| 8, | |
| 4, | |
| 8, | |
| 4, | |
| 16, | |
| 4, | |
| 32, | |
| 4, | |
| 64, | |
| 4 | |
| ], | |
| "moe_min_capacity": 4, | |
| "moe_noisy_gate_policy": null, | |
| "moe_output_dropout_prob": 0.0, | |
| "moe_use_residual": true, | |
| "moe_use_rts": true, | |
| "num_experts_per_tok": 1, | |
| "num_hidden_layers": 23, | |
| "num_kv_heads": 8, | |
| "num_query_heads": 16, | |
| "num_species": 99700, | |
| "pad_token_id": 0, | |
| "prepend_species_token": true, | |
| "rope_theta": 500000.0, | |
| "router_aux_loss_coef": 0.001, | |
| "species_emb_dim": 32, | |
| "species_encoder_embedding_dim": 32, | |
| "species_encoder_pooling_strategy": "eos", | |
| "species_encoder_repo": null, | |
| "species_encoder_revision": null, | |
| "species_encoder_subfolder": "species_encoder", | |
| "species_vocab_sha256": "7f46fd7aebc6980c7cb4d5b3505009bee87b3f93e1ee7fdbc4146bf6b56e61e1", | |
| "tie_word_embeddings": true, | |
| "tokenizer_class": "MicroGlotTokenizerFast", | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.51.3", | |
| "use_flash_attention": true, | |
| "use_swiglu": true, | |
| "vocab_size": 8192 | |
| } | |