Instructions to use MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer

SGLang

How to use MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer 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 "MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer with Docker Model Runner:
```
docker model run hf.co/MTSmash/EvaGPT-German-2B-Q11-Pretrained-own-Tokenizer
```

MTSmash commited on 2 days ago

Commit

47da018

verified ·

1 Parent(s): 50824c9

Upload model

Browse files

Files changed (3) hide show

config.json +69 -0
generation_config.json +11 -0
model.safetensors +3 -0

config.json ADDED Viewed

	@@ -0,0 +1,69 @@

+{
+  "architectures": [
+    "EvaGptForCausalLM"
+  ],
+  "attention_bias": true,
+  "attention_dropout": 0.0,
+  "bos_token_id": 3,
+  "dtype": "bfloat16",
+  "eos_token_id": 3,
+  "experts_per_token": 4,
+  "head_dim": 64,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initial_context_length": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 8192,
+  "layer_types": [
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 131072,
+  "model_type": "eva_gpt",
+  "num_attention_heads": 32,
+  "num_experts_per_tok": 1,
+  "num_hidden_layers": 16,
+  "num_key_value_heads": 8,
+  "num_local_experts": 1,
+  "output_router_logits": false,
+  "pad_token_id": 4,
+  "quantization_config": {
+    "modules_to_not_convert": [
+      "model.layers.*.self_attn",
+      "model.layers.*.mlp.router",
+      "model.embed_tokens",
+      "lm_head"
+    ],
+    "quant_method": "mxfp4"
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_parameters": {
+    "beta_fast": 32.0,
+    "beta_slow": 1.0,
+    "factor": 32.0,
+    "original_max_position_embeddings": 4096,
+    "rope_theta": 150000,
+    "rope_type": "yarn"
+  },
+  "router_aux_loss_coef": 0.9,
+  "sliding_window": 128,
+  "swiglu_limit": 7.0,
+  "tie_word_embeddings": false,
+  "transformers_version": "5.0.0.dev0",
+  "use_cache": true,
+  "vocab_size": 200022
+}

generation_config.json ADDED Viewed

	@@ -0,0 +1,11 @@

+{
+  "_from_model_config": true,
+  "bos_token_id": 3,
+  "eos_token_id": 3,
+  "max_new_tokens": 2336,
+  "min_new_tokens": 2,
+  "num_beams": 1,
+  "pad_token_id": 4,
+  "repetition_penalty": 1.1,
+  "transformers_version": "5.0.0.dev0"
+}

model.safetensors ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:81119a9fa5267b45381f9adeb982b2f63c7b53d5ba47a021b3730caa9e6fec77
+size 3585723880