fix extra name

.github/workflows/test_cli_cpu_llama_cpp.yaml

@@ -42,7 +42,7 @@ jobs:
         run: |
           pip install --upgrade pip
           pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu
-          pip install -e .[testing,lamma-cpp]
+          pip install -e .[testing,llama-cpp]
 
       - name: Run tests
         run: pytest -s -k "llama_cpp"

setup.py

@@ -75,10 +75,10 @@
     "neural-compressor": [f"optimum[neural-compressor]>={MIN_OPTIMUM_VERSION}"],
     "torch-ort": ["torch-ort", "onnxruntime-training", f"optimum>={MIN_OPTIMUM_VERSION}"],
     # other backends
+    "llama-cpp": ["llama-cpp-python"],
     "llm-swarm": ["llm-swarm"],
     "py-txi": ["py-txi"],
     "vllm": ["vllm"],
-    "lamma-cpp": ["llama-cpp-python"],
     # optional dependencies
     "autoawq": [AUTOAWQ],
     "auto-gptq": ["optimum", AUTOGPTQ],

tests/configs/cpu_inference_llama_cpp_embedding.yaml

@@ -1,23 +1,21 @@
 defaults:
-  - _base_
-  - _inference_
-  - _self_
+  # order of inheritance, last one overrides previous ones
+  - _base_ # inherits from base config
+  - _cpu_ # inherits from cpu config
+  - _inference_ # inherits from inference config
+  - _self_ # hydra 1.1 compatibility
   - override backend: llama_cpp
 
 name: inference_llama_cpp_embedding
 
 backend:
   model: nomic-ai/nomic-embed-text-v1.5-GGUF
-  task: feature-extraction
   filename: nomic-embed-text-v1.5.Q4_0.gguf
 
 scenario:
   input_shapes:
     batch_size: 1
     sequence_length: 256
-    vocab_size: 30000
-    type_vocab_size: 1
-    max_position_embeddings: 512
   generate_kwargs:
     max_new_tokens: 100
-    min_new_tokens: 100
+    min_new_tokens: 100

tests/configs/cpu_inference_llama_cpp_text_generation.yaml

@@ -1,21 +1,21 @@
 defaults:
-  - _base_
-  - _inference_
-  - _self_
+  # order of inheritance, last one overrides previous ones
+  - _base_ # inherits from base config
+  - _cpu_ # inherits from cpu config
+  - _inference_ # inherits from inference config
+  - _self_ # hydra 1.1 compatibility
   - override backend: llama_cpp
 
-name: inference_llama_cpp_text_generation
+name: cpu_inference_llama_cpp_text_generation
 
 backend:
   model: TheBloke/TinyLlama-1.1B-Chat-v1.0-GGUF
-  task: text-generation
   filename: tinyllama-1.1b-chat-v1.0.Q4_0.gguf
 
 scenario:
   input_shapes:
     batch_size: 1
     sequence_length: 256
-    vocab_size: 32000
   generate_kwargs:
     max_new_tokens: 100
-    min_new_tokens: 100
+    min_new_tokens: 100