Text Generation
Transformers
Safetensors
English
slmoe
causal-lm
base-model
mixture-of-experts
sequence-routing
custom-code
trust-remote-code
custom_code
Instructions to use Banaxi-Tech/slmoe-test with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Banaxi-Tech/slmoe-test with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="Banaxi-Tech/slmoe-test", trust_remote_code=True)# Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("Banaxi-Tech/slmoe-test", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use Banaxi-Tech/slmoe-test with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Banaxi-Tech/slmoe-test" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Banaxi-Tech/slmoe-test", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/Banaxi-Tech/slmoe-test
- SGLang
How to use Banaxi-Tech/slmoe-test with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "Banaxi-Tech/slmoe-test" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Banaxi-Tech/slmoe-test", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "Banaxi-Tech/slmoe-test" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Banaxi-Tech/slmoe-test", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use Banaxi-Tech/slmoe-test with Docker Model Runner:
docker model run hf.co/Banaxi-Tech/slmoe-test
File size: 2,482 Bytes
4886f06 e6997cf 4886f06 e6997cf 4886f06 e6997cf 4886f06 e6997cf 4886f06 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 | {
"parameters": 25449472,
"active_parameters": 7902208,
"architecture": {
"vocab_size": 8192,
"hidden_size": 256,
"num_hidden_layers": 8,
"num_attention_heads": 8,
"num_key_value_heads": 2,
"head_dim": 32,
"num_experts": 64,
"num_experts_per_sequence": 13,
"expert_intermediate_size": 56,
"router_prefix_length": 32,
"router_jitter_noise": 0.01,
"router_aux_loss_coeff": 0.01,
"router_z_loss_coeff": 0.001,
"expert_output_scale": 3.605551275463989,
"max_position_embeddings": 4096,
"rope_theta": 100000.0,
"rms_norm_eps": 1e-06,
"initializer_range": 0.02,
"use_cache": true,
"transformers_version": "5.15.0",
"architectures": null,
"output_hidden_states": false,
"return_dict": true,
"dtype": null,
"chunk_size_feed_forward": 0,
"is_encoder_decoder": false,
"id2label": {
"0": "LABEL_0",
"1": "LABEL_1"
},
"label2id": {
"LABEL_0": 0,
"LABEL_1": 1
},
"problem_type": null,
"_name_or_path": "",
"tie_word_embeddings": true,
"bos_token_id": 1,
"eos_token_id": 2,
"pad_token_id": 0,
"unk_token_id": 3,
"model_type": "slmoe",
"output_attentions": false
},
"training_percent": 100,
"step": 28610,
"total_steps": 28610,
"tokens_seen": 59999518720,
"target_tokens": 60000000000,
"trained_token_target": 59999518720,
"tokens_per_step": 2097152,
"tokens_per_micro_batch": 1048576,
"local_batch": 64,
"effective_global_batch": 512,
"gradient_accumulation_steps": 2,
"world_size": 4,
"gpu_name": "NVIDIA H200",
"optimizer": "AdamW",
"peak_lr": 0.003,
"architecture_revision": "bd0ade7689958997e1889c641a782a69c0bea918",
"tokenizer_repo": "BananaMind/BananaMind-2-Mini",
"tokenizer_revision": "6400d0a6dcbe52f3c725291cb97c079c591c05a4",
"dataset_revisions": {
"fineweb_hq": "e58199cdd52438d94405df1a4d8630cc5f13bf84",
"fineweb_edu": "87f09149ef4734204d70ed1d046ddc9ca3f2b8f9",
"dclm": "a3b142c183aebe5af344955ae20836eb34dcf69b",
"cosmopedia_v2": "3ba9d605774198c5868892d7a8deda78031a781f",
"finemath": "e92b25a616738fe95dc186b64dfb19f9c8525594",
"npset2": "222b911cd9119624dd65ab084d69542f6e04631e"
},
"source_tokens": {
"fineweb_hq": 16919822336,
"fineweb_edu": 16949182464,
"dclm": 9330229248,
"cosmopedia_v2": 7499415552,
"finemath": 7470055424,
"npset2": 1830813696
},
"elapsed_seconds": 15931.297726869583
}
|