Switch to optimum-intel internal testing models

tests/python_tests/data/models.py

@@ -8,7 +8,7 @@
 
 def get_models_list():
     model_ids = [
-        "katuni4ka/tiny-random-phi3",
+        "optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM",
     ]
     if pytest.selected_model_ids:
         model_ids = [model_id for model_id in model_ids if model_id in pytest.selected_model_ids.split(' ')]

tests/python_tests/data/tokenizer_configs.py

@@ -670,7 +670,7 @@ def get_tokenizer_configs():
             "unk_token": "<unk>",
             "chat_template": "{% for message in messages %}{% if message['role'] == 'system' %}{{'<|system|>'+ '\n' + message['content'] + '\n'}}{% elif message['role'] == 'user' %}{{'<|user|>' + '\n' + message['content'] + '\n'}}{% elif message['role'] == 'assistant' %}{{'<|assistant|>' + '\n' + message['content'] + '<|endoftext|>' + ('' if loop.last else '\n')}}{% endif %}{% endfor %}"
         },
-        "katuni4ka/tiny-random-phi3": {
+        "optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM": {
             "bos_token": "<s>",
             "eos_token": "<|endoftext|>",
             "pad_token": "<|endoftext|>",
@@ -712,7 +712,7 @@ def get_tokenizer_configs():
             "unk_token": "<unk>",
             "chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '### User:\n' + message['content'] }}\n{% elif message['role'] == 'system' %}\n{{ '### System:\n' + message['content'] }}\n{% elif message['role'] == 'assistant' %}\n{{ '### Assistant:\n'  + message['content'] }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '### Assistant:' }}\n{% endif %}\n{% endfor %}"
         },
-        "katuni4ka/tiny-random-minicpm": {
+        "optimum-intel-internal-testing/tiny-random-minicpm": {
             "bos_token": "<s>",
             "eos_token": "</s>",
             "pad_token": None,

tests/python_tests/samples/conftest.py

@@ -97,7 +97,7 @@
         "convert_args": []
     },
     "tiny-random-minicpmv-2_6": {
-        "name": "katuni4ka/tiny-random-minicpmv-2_6",
+        "name": "optimum-intel-internal-testing/tiny-random-minicpmv-2_6",
         "convert_args": ['--trust-remote-code', "--task", "image-text-to-text"]
     },
     "InternVL2-1B": {
@@ -121,15 +121,15 @@
         "convert_args": ["--task", "text-generation-with-past", "--weight-format", "int8"]
     },
     "tiny-random-latent-consistency": {
-        "name": "echarlaix/tiny-random-latent-consistency",
+        "name": "optimum-intel-internal-testing/tiny-random-latent-consistency",
         "convert_args": ['--trust-remote-code', '--weight-format', 'fp16']
     },
     "tiny-random-latent-consistency-lora": {
         "name": "katuni4ka/tiny-random-latent-consistency-lora",
         "convert_args": []
     },
     "tiny-random-llava": {
-        "name": "katuni4ka/tiny-random-llava",
+        "name": "optimum-intel-internal-testing/tiny-random-llava",
         "convert_args": ["--trust-remote-code", "--task", "image-text-to-text"]
     },
     "bge-small-en-v1.5": {

tests/python_tests/test_llm_pipeline.py

@@ -99,7 +99,7 @@ def test_batch_string_inputs(model_id, generation_config_dict, prompts, pipeline
 
 @pytest.mark.precommit
 def test_batch_size_switch():
-    model_id = 'katuni4ka/tiny-random-phi3'
+    model_id = 'optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM'
     _, _, models_path = download_and_convert_model(model_id)
     ov_pipe = create_ov_pipeline(models_path)
 
@@ -110,7 +110,7 @@ def test_batch_size_switch():
 
 @pytest.mark.precommit
 def test_empty_encoded_inputs_throw():
-    model_id = 'katuni4ka/tiny-random-phi3'
+    model_id = 'optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM'
     _, _, models_path = download_and_convert_model(model_id)
     ov_pipe = create_ov_pipeline(models_path)
 
@@ -645,37 +645,43 @@ def test_pipeline_validates_generation_config(model_id):
 # Work with Unicode in Python API
 #
 
+# Model, prompt and max_new_tokens that generates unfinished utf-8 string.
+UNICODE_PYBIND_DECODING_TEST_CASES: list[tuple[str, str, int]] = [
+    ("optimum-intel-internal-testing/tiny-random-PhiForCausalLM", ",", 3)
+]
+
+
 @pytest.mark.precommit
-@pytest.mark.parametrize("model_id", get_models_list())
-def test_unicode_pybind_decoding_one_string(model_id):
+@pytest.mark.parametrize("model_id,prompt,max_new_tokens", UNICODE_PYBIND_DECODING_TEST_CASES)
+def test_unicode_pybind_decoding_one_string(model_id: str, prompt: str, max_new_tokens: int):
     # On this model this prompt generates unfinished utf string.
     # Test that pybind will not fail.
     _, _, models_path = download_and_convert_model(model_id)
     ov_pipe = create_ov_pipeline(models_path)
-    res_str = ov_pipe.generate(',', max_new_tokens=4, apply_chat_template=False)
+    res_str = ov_pipe.generate(prompt, max_new_tokens=max_new_tokens, apply_chat_template=False)
     assert '�' == res_str[-1]
 
 
 @pytest.mark.precommit
-@pytest.mark.parametrize("model_id", get_models_list())
-def test_unicode_pybind_decoding_batched(model_id):
+@pytest.mark.parametrize("model_id,prompt,max_new_tokens", UNICODE_PYBIND_DECODING_TEST_CASES)
+def test_unicode_pybind_decoding_batched(model_id: str, prompt: str, max_new_tokens: int):
     # On this model this prompt generates unfinished utf string.
     # Test that pybind will not fail.
     _, _, models_path = download_and_convert_model(model_id)
     ov_pipe = create_ov_pipeline(models_path)
-    res_str = ov_pipe.generate([","], max_new_tokens=4, apply_chat_template=False)
+    res_str = ov_pipe.generate([prompt], max_new_tokens=max_new_tokens, apply_chat_template=False)
     assert '�' == res_str.texts[0][-1]
 
 
 @pytest.mark.precommit
-@pytest.mark.parametrize("model_id", get_models_list())
-def test_unicode_pybind_decoding_one_string_streamer(model_id):
+@pytest.mark.parametrize("model_id,prompt,max_new_tokens", UNICODE_PYBIND_DECODING_TEST_CASES)
+def test_unicode_pybind_decoding_one_string_streamer(model_id: str, prompt: str, max_new_tokens: int):
     # On this model this prompt generates unfinished utf-8 string
     # and streams it. Test that pybind will not fail while we pass string to python.
     _, _, models_path = download_and_convert_model(model_id)
     ov_pipe = create_ov_pipeline(models_path)
     res_str = []
-    ov_pipe.generate(",", max_new_tokens=4, apply_chat_template=False, streamer=lambda x: res_str.append(x))
+    ov_pipe.generate(prompt, max_new_tokens=max_new_tokens, apply_chat_template=False, streamer=lambda x: res_str.append(x))
     assert '�' == ''.join(res_str)[-1]
 
 #
@@ -696,7 +702,7 @@ def run_perf_metrics_collection(model_id, generation_config_dict: dict, prompt:
 def test_perf_metrics(generation_config, prompt):
     import time
     start_time = time.perf_counter()
-    model_id = 'katuni4ka/tiny-random-gemma2'
+    model_id = 'optimum-intel-internal-testing/tiny-random-gemma2'
     perf_metrics = run_perf_metrics_collection(model_id, generation_config, prompt)
     total_time = (time.perf_counter() - start_time) * 1000
 
@@ -778,7 +784,7 @@ class Person(BaseModel):
         city: Literal["Dublin", "Dubai", "Munich"]
     generation_config.update(dict(structured_output_config=ov_genai.StructuredOutputConfig(json_schema=json.dumps(Person.model_json_schema()))))
 
-    model_id = 'katuni4ka/tiny-random-gemma2'
+    model_id = 'optimum-intel-internal-testing/tiny-random-gemma2'
     _, _, models_path = download_and_convert_model(model_id)
     ov_pipe = create_ov_pipeline(models_path)
     perf_metrics = ov_pipe.generate([prompt], **generation_config).perf_metrics

tests/python_tests/test_sampling.py

@@ -29,7 +29,7 @@
                             #   "echo_with_generation",
                               ])
 def test_basic_stop_criteria(generation_config, prompt):
-    model_id : str = "katuni4ka/tiny-random-phi3"
+    model_id : str = "optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM"
     generate_and_compare(model_id, [prompt], generation_config)
 
 
@@ -71,7 +71,7 @@ def test_stop_strings(generation_config, model_id, pipeline_type):
     'I have an interview about product speccing with the company Weekend Health. Give me an example of a question they might ask with regards about a new feature'
 ])
 def test_greedy(generation_config, prompt):
-    model_id : str = "katuni4ka/tiny-random-phi3"
+    model_id : str = "optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM"
     prompt = prompt.decode('unicode_escape') if isinstance(prompt, bytes) else prompt
 
     generate_and_compare(model=model_id, 

tests/python_tests/test_stateful_speculative_decoding.py

@@ -71,7 +71,7 @@ def test_string_inputs(main_model, main_device, draft_model, draft_device, promp
 def test_perf_metrics():
     import time
     start_time = time.perf_counter()
-    model_id = 'katuni4ka/tiny-random-gemma2'
+    model_id = 'optimum-intel-internal-testing/tiny-random-gemma2'
     _, _, model_path = download_and_convert_model(model_id)
 
     # Create OpenVINO GenAI pipeline:

tests/python_tests/test_structured_output.py

@@ -35,7 +35,7 @@ class RESTAPIResponse(BaseModel):
 
 structured_id_models = [
     "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
-    "katuni4ka/tiny-random-phi3",
+    "optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM",
 ]
 
 

tests/python_tests/test_tokenizer.py

@@ -308,7 +308,7 @@ def test_set_chat_template(ov_hf_tokenizers):
 @pytest.mark.parametrize(
     "ov_hf_tokenizers",
     [
-        "katuni4ka/tiny-random-phi3",
+        "optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM",
         "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
         # ("black-forest-labs/FLUX.1-dev", dict(subfolder="tokenizer")),  # FLUX.1-dev has tokenizer in subfolder
     ],
@@ -420,10 +420,10 @@ def hf_ov_genai_models(request, tmp_path_factory):
 @pytest.mark.parametrize(
     "hf_ov_genai_models",
     [
-        ("katuni4ka/tiny-random-phi3", {"padding_side": None}),
+        ("optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM", {"padding_side": None}),
         ("TinyLlama/TinyLlama-1.1B-Chat-v1.0", {"padding_side": None}),
-        ("katuni4ka/tiny-random-llava-next", {"padding_side": "right"}),
-        ("katuni4ka/tiny-random-llava-next", {"padding_side": "left"}),
+        ("optimum-intel-internal-testing/tiny-random-llava-next", {"padding_side": "right"}),
+        ("optimum-intel-internal-testing/tiny-random-llava-next", {"padding_side": "left"}),
         (
             "BAAI/bge-small-en-v1.5",
             {"padding_side": None},
@@ -485,8 +485,8 @@ def test_padding(
 base_models_for_paired_input_test = [
     ("answerdotai/ModernBERT-base", {"padding_side": None}),
     ("TinyLlama/TinyLlama-1.1B-Chat-v1.0", {"padding_side": None}),
-    ("katuni4ka/tiny-random-llava-next", {"padding_side": "right"}),
-    ("katuni4ka/tiny-random-llava-next", {"padding_side": "left"}),
+    ("optimum-intel-internal-testing/tiny-random-llava-next", {"padding_side": "right"}),
+    ("optimum-intel-internal-testing/tiny-random-llava-next", {"padding_side": "left"}),
 ]
 
 def make_model_params():