| Epoch | \n", "Training Loss | \n", "Validation Loss | \n", "Rouge1 | \n", "Rouge2 | \n", "Rougel | \n", "Rougelsum | \n", "Gen Len | \n", "
|---|---|---|---|---|---|---|---|
| 1 | \n", "1.439000 | \n", "1.381694 | \n", "47.236000 | \n", "23.568400 | \n", "39.745000 | \n", "43.495300 | \n", "17.134310 | \n", "
| 2 | \n", "1.355200 | \n", "1.371634 | \n", "46.913200 | \n", "23.134100 | \n", "39.428800 | \n", "43.234700 | \n", "17.416361 | \n", "
| 3 | \n", "1.280100 | \n", "1.370761 | \n", "47.476000 | \n", "23.812500 | \n", "40.079200 | \n", "43.582600 | \n", "17.134310 | \n", "
| 4 | \n", "1.231200 | \n", "1.372878 | \n", "47.429500 | \n", "23.850800 | \n", "40.067500 | \n", "43.599700 | \n", "17.203907 | \n", "
| 5 | \n", "1.193500 | \n", "1.376878 | \n", "47.378900 | \n", "23.925500 | \n", "40.027400 | \n", "43.725000 | \n", "17.221001 | \n", "
"
]
},
"metadata": {}
},
{
"output_type": "execute_result",
"data": {
"text/plain": [
"TrainOutput(global_step=9210, training_loss=1.3033124296206993, metrics={'train_runtime': 3255.6159, 'train_samples_per_second': 22.626, 'train_steps_per_second': 2.829, 'total_flos': 5.043922658131968e+16, 'train_loss': 1.3033124296206993, 'epoch': 5.0})"
]
},
"metadata": {},
"execution_count": 15
}
]
},
{
"cell_type": "code",
"source": [
"trainer.evaluate()"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 211
},
"id": "xoqs0hXRoj_w",
"outputId": "eeb09c6e-cee3-443d-ab99-41af0749a848"
},
"execution_count": 16,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮\n",
"│ /usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py:261 in │\n",
"│ hf_raise_for_status │\n",
"│ │\n",
"│ 258 │ </Tip> │\n",
"│ 259 │ \"\"\" │\n",
"│ 260 │ try: │\n",
"│ ❱ 261 │ │ response.raise_for_status() │\n",
"│ 262 │ except HTTPError as e: │\n",
"│ 263 │ │ error_code = response.headers.get(\"X-Error-Code\") │\n",
"│ 264 │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/requests/models.py:960 in raise_for_status │\n",
"│ │\n",
"│ 957 │ │ │ http_error_msg = u'%s Server Error: %s for url: %s' % (self.status_code, rea │\n",
"│ 958 │ │ │\n",
"│ 959 │ │ if http_error_msg: │\n",
"│ ❱ 960 │ │ │ raise HTTPError(http_error_msg, response=self) │\n",
"│ 961 │ │\n",
"│ 962 │ def close(self): │\n",
"│ 963 │ │ \"\"\"Releases the connection back to the pool. Once this method has been │\n",
"╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\n",
"HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/api/repos/create\n",
"\n",
"The above exception was the direct cause of the following exception:\n",
"\n",
"╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮\n",
"│ /usr/local/lib/python3.10/dist-packages/huggingface_hub/hf_api.py:2308 in create_repo │\n",
"│ │\n",
"│ 2305 │ │ r = get_session().post(path, headers=headers, json=json) │\n",
"│ 2306 │ │ │\n",
"│ 2307 │ │ try: │\n",
"│ ❱ 2308 │ │ │ hf_raise_for_status(r) │\n",
"│ 2309 │ │ except HTTPError as err: │\n",
"│ 2310 │ │ │ if exist_ok and err.response.status_code == 409: │\n",
"│ 2311 │ │ │ │ # Repo already exists and `exist_ok=True` │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py:303 in │\n",
"│ hf_raise_for_status │\n",
"│ │\n",
"│ 300 │ │ │\n",
"│ 301 │ │ # Convert `HTTPError` into a `HfHubHTTPError` to display request information │\n",
"│ 302 │ │ # as well (request id and/or server error message) │\n",
"│ ❱ 303 │ │ raise HfHubHTTPError(str(e), response=response) from e │\n",
"│ 304 │\n",
"│ 305 │\n",
"│ 306 def _format_error_message(message: str, request_id: Optional[str], server_message: Optio │\n",
"╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\n",
"HfHubHTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/api/repos/create (Request ID: \n",
"Root=1-64a9f7dc-08f648210c1658f20642bcba;3e3a5cd7-ea5a-4a1b-b450-a06ae15315ec)\n",
"\n",
"You don't have the rights to create a model under this namespace\n",
"\n",
"During handling of the above exception, another exception occurred:\n",
"\n",
"╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮\n",
"│ /usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py:261 in │\n",
"│ hf_raise_for_status │\n",
"│ │\n",
"│ 258 │ </Tip> │\n",
"│ 259 │ \"\"\" │\n",
"│ 260 │ try: │\n",
"│ ❱ 261 │ │ response.raise_for_status() │\n",
"│ 262 │ except HTTPError as e: │\n",
"│ 263 │ │ error_code = response.headers.get(\"X-Error-Code\") │\n",
"│ 264 │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/requests/models.py:960 in raise_for_status │\n",
"│ │\n",
"│ 957 │ │ │ http_error_msg = u'%s Server Error: %s for url: %s' % (self.status_code, rea │\n",
"│ 958 │ │ │\n",
"│ 959 │ │ if http_error_msg: │\n",
"│ ❱ 960 │ │ │ raise HTTPError(http_error_msg, response=self) │\n",
"│ 961 │ │\n",
"│ 962 │ def close(self): │\n",
"│ 963 │ │ \"\"\"Releases the connection back to the pool. Once this method has been │\n",
"╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\n",
"HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/api/models/JeffGuo/flan-t5-base-samsum\n",
"\n",
"The above exception was the direct cause of the following exception:\n",
"\n",
"╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮\n",
"│ in <cell line: 5>:5 │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/transformers/trainer.py:3716 in push_to_hub │\n",
"│ │\n",
"│ 3713 │ │ # If a user calls manually `push_to_hub` with `self.args.push_to_hub = False`, w │\n",
"│ 3714 │ │ # it might fail. │\n",
"│ 3715 │ │ if not hasattr(self, \"repo\"): │\n",
"│ ❱ 3716 │ │ │ self.init_git_repo() │\n",
"│ 3717 │ │ │\n",
"│ 3718 │ │ model_name = kwargs.pop(\"model_name\", None) │\n",
"│ 3719 │ │ if model_name is None and self.args.should_save: │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/transformers/trainer.py:3569 in init_git_repo │\n",
"│ │\n",
"│ 3566 │ │ │ repo_name = get_full_repo_name(repo_name, token=self.args.hub_token) │\n",
"│ 3567 │ │ │\n",
"│ 3568 │ │ # Make sure the repo exists. │\n",
"│ ❱ 3569 │ │ create_repo(repo_name, token=self.args.hub_token, private=self.args.hub_private_ │\n",
"│ 3570 │ │ try: │\n",
"│ 3571 │ │ │ self.repo = Repository(self.args.output_dir, clone_from=repo_name, token=sel │\n",
"│ 3572 │ │ except EnvironmentError: │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py:118 in _inner_fn │\n",
"│ │\n",
"│ 115 │ │ if check_use_auth_token: │\n",
"│ 116 │ │ │ kwargs = smoothly_deprecate_use_auth_token(fn_name=fn.__name__, has_token=ha │\n",
"│ 117 │ │ │\n",
"│ ❱ 118 │ │ return fn(*args, **kwargs) │\n",
"│ 119 │ │\n",
"│ 120 │ return _inner_fn # type: ignore │\n",
"│ 121 │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/huggingface_hub/hf_api.py:2316 in create_repo │\n",
"│ │\n",
"│ 2313 │ │ │ elif exist_ok and err.response.status_code == 403: │\n",
"│ 2314 │ │ │ │ # No write permission on the namespace but repo might already exist │\n",
"│ 2315 │ │ │ │ try: │\n",
"│ ❱ 2316 │ │ │ │ │ self.repo_info(repo_id=repo_id, repo_type=repo_type, token=token) │\n",
"│ 2317 │ │ │ │ │ if repo_type is None or repo_type == REPO_TYPE_MODEL: │\n",
"│ 2318 │ │ │ │ │ │ return RepoUrl(f\"{self.endpoint}/{repo_id}\") │\n",
"│ 2319 │ │ │ │ │ return RepoUrl(f\"{self.endpoint}/{repo_type}/{repo_id}\") │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py:118 in _inner_fn │\n",
"│ │\n",
"│ 115 │ │ if check_use_auth_token: │\n",
"│ 116 │ │ │ kwargs = smoothly_deprecate_use_auth_token(fn_name=fn.__name__, has_token=ha │\n",
"│ 117 │ │ │\n",
"│ ❱ 118 │ │ return fn(*args, **kwargs) │\n",
"│ 119 │ │\n",
"│ 120 │ return _inner_fn # type: ignore │\n",
"│ 121 │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/huggingface_hub/hf_api.py:1868 in repo_info │\n",
"│ │\n",
"│ 1865 │ │ │ method = self.space_info # type: ignore │\n",
"│ 1866 │ │ else: │\n",
"│ 1867 │ │ │ raise ValueError(\"Unsupported repo type.\") │\n",
"│ ❱ 1868 │ │ return method( │\n",
"│ 1869 │ │ │ repo_id, │\n",
"│ 1870 │ │ │ revision=revision, │\n",
"│ 1871 │ │ │ token=token, │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py:118 in _inner_fn │\n",
"│ │\n",
"│ 115 │ │ if check_use_auth_token: │\n",
"│ 116 │ │ │ kwargs = smoothly_deprecate_use_auth_token(fn_name=fn.__name__, has_token=ha │\n",
"│ 117 │ │ │\n",
"│ ❱ 118 │ │ return fn(*args, **kwargs) │\n",
"│ 119 │ │\n",
"│ 120 │ return _inner_fn # type: ignore │\n",
"│ 121 │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/huggingface_hub/hf_api.py:1678 in model_info │\n",
"│ │\n",
"│ 1675 │ │ if files_metadata: │\n",
"│ 1676 │ │ │ params[\"blobs\"] = True │\n",
"│ 1677 │ │ r = get_session().get(path, headers=headers, timeout=timeout, params=params) │\n",
"│ ❱ 1678 │ │ hf_raise_for_status(r) │\n",
"│ 1679 │ │ d = r.json() │\n",
"│ 1680 │ │ return ModelInfo(**d) │\n",
"│ 1681 │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py:293 in │\n",
"│ hf_raise_for_status │\n",
"│ │\n",
"│ 290 │ │ │ │ \" `repo_type`.\\nIf you are trying to access a private or gated repo,\" │\n",
"│ 291 │ │ │ │ \" make sure you are authenticated.\" │\n",
"│ 292 │ │ │ ) │\n",
"│ ❱ 293 │ │ │ raise RepositoryNotFoundError(message, response) from e │\n",
"│ 294 │ │ │\n",
"│ 295 │ │ elif response.status_code == 400: │\n",
"│ 296 │ │ │ message = ( │\n",
"╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\n",
"RepositoryNotFoundError: 404 Client Error. (Request ID: \n",
"Root=1-64a9f7dd-665988123c3b7a88361ad5a9;8254f7f7-f061-405e-aa6b-5911f96d65dc)\n",
"\n",
"Repository Not Found for url: https://huggingface.co/api/models/JeffGuo/flan-t5-base-samsum.\n",
"Please make sure you specified the correct `repo_id` and `repo_type`.\n",
"If you are trying to access a private or gated repo, make sure you are authenticated.\n",
"
\n"
]
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"from transformers import pipeline\n",
"from random import randrange\n",
"\n",
"# load model and tokenizer from huggingface hub with pipeline\n",
"summarizer = pipeline(\"summarization\", model=\"/content/flan-t5-base-samsum/checkpoint-5526\", device=0)\n",
"#summarizer = pipeline(\"summarization\", model=\"/content/flan-t5-base-samsum/checkpoint-5526/content/t5-small-finetuned-xsum/checkpoint-12500\", device=0)\n",
"\n",
"# select a random test sample\n",
"sample = dataset['test'][randrange(len(dataset[\"test\"]))]\n",
"print(f\"dialogue: \\n{sample['dialogue']}\\n---------------\")\n",
"\n",
"# summarize dialogue\n",
"res = summarizer(sample[\"dialogue\"])\n",
"\n",
"print(f\"flan-t5-base summary:\\n{res[0]['summary_text']}\")"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 666
},
"id": "kcsZspp_ovD3",
"outputId": "aea9b08c-16ca-40bf-f944-2639f784ec15"
},
"execution_count": 29,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"\u001b[31m╭─\u001b[0m\u001b[31m──────────────────────────────\u001b[0m\u001b[31m \u001b[0m\u001b[1;31mTraceback \u001b[0m\u001b[1;2;31m(most recent call last)\u001b[0m\u001b[31m \u001b[0m\u001b[31m───────────────────────────────\u001b[0m\u001b[31m─╮\u001b[0m\n",
"\u001b[31m│\u001b[0m in \u001b[92m╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮\n",
"│ in <cell line: 5>:5 │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/transformers/pipelines/__init__.py:885 in pipeline │\n",
"│ │\n",
"│ 882 │ │ │ │ tokenizer_kwargs = model_kwargs.copy() │\n",
"│ 883 │ │ │ │ tokenizer_kwargs.pop(\"torch_dtype\", None) │\n",
"│ 884 │ │ │ │\n",
"│ ❱ 885 │ │ │ tokenizer = AutoTokenizer.from_pretrained( │\n",
"│ 886 │ │ │ │ tokenizer_identifier, use_fast=use_fast, _from_pipeline=task, **hub_kwar │\n",
"│ 887 │ │ │ ) │\n",
"│ 888 │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/transformers/models/auto/tokenization_auto.py:709 in │\n",
"│ from_pretrained │\n",
"│ │\n",
"│ 706 │ │ if model_type is not None: │\n",
"│ 707 │ │ │ tokenizer_class_py, tokenizer_class_fast = TOKENIZER_MAPPING[type(config)] │\n",
"│ 708 │ │ │ if tokenizer_class_fast and (use_fast or tokenizer_class_py is None): │\n",
"│ ❱ 709 │ │ │ │ return tokenizer_class_fast.from_pretrained(pretrained_model_name_or_pat │\n",
"│ 710 │ │ │ else: │\n",
"│ 711 │ │ │ │ if tokenizer_class_py is not None: │\n",
"│ 712 │ │ │ │ │ return tokenizer_class_py.from_pretrained(pretrained_model_name_or_p │\n",
"│ │\n",
"│ /usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils_base.py:1809 in │\n",
"│ from_pretrained │\n",
"│ │\n",
"│ 1806 │ │ │ ) │\n",
"│ 1807 │ │ │\n",
"│ 1808 │ │ if all(full_file_name is None for full_file_name in resolved_vocab_files.values( │\n",
"│ ❱ 1809 │ │ │ raise EnvironmentError( │\n",
"│ 1810 │ │ │ │ f\"Can't load tokenizer for '{pretrained_model_name_or_path}'. If you wer │\n",
"│ 1811 │ │ │ │ \"'https://huggingface.co/models', make sure you don't have a local direc │\n",
"│ 1812 │ │ │ │ f\"Otherwise, make sure '{pretrained_model_name_or_path}' is the correct │\n",
"╰──────────────────────────────────────────────────────────────────────────────────────────────────╯\n",
"OSError: Can't load tokenizer for '/content/flan-t5-base-samsum/checkpoint-5526'. If you were trying to load it \n",
"from 'https://huggingface.co/models', make sure you don't have a local directory with the same name. Otherwise, \n",
"make sure '/content/flan-t5-base-samsum/checkpoint-5526' is the correct path to a directory containing all relevant\n",
"files for a T5TokenizerFast tokenizer.\n",
"
\n"
]
},
"metadata": {}
}
]
}
]
}