leadmaister commited on
Commit
d7acaf3
1 Parent(s): d0259c9

Update config.gradio.yaml

Browse files
Files changed (1) hide show
  1. config.gradio.yaml +8 -8
config.gradio.yaml CHANGED
@@ -1,14 +1,14 @@
1
  openai:
2
- key: gradio # "gradio" (set when request) or your_personal_key
3
  huggingface:
4
- token: # required: huggingface token @ https://huggingface.co/settings/tokens
5
  dev: false
6
  debug: true
7
  log_file: logs/debug_TIMESTAMP.log
8
  model: text-davinci-003 # text-davinci-003
9
  use_completion: true
10
  inference_mode: hybrid # local, huggingface or hybrid
11
- local_deployment: standard # minimal, standard or full
12
  num_candidate_models: 5
13
  max_description_length: 100
14
  proxy:
@@ -17,18 +17,18 @@ logit_bias:
17
  choose_model: 5
18
  tprompt:
19
  parse_task: >-
20
- #1 Task Planning Stage: The AI assistant can parse user input to several tasks: [{"task": task, "id": task_id, "dep": dependency_task_id, "args": {"text": text or <GENERATED>-dep_id, "image": image_url or <GENERATED>-dep_id, "audio": audio_url or <GENERATED>-dep_id}}]. The special tag "<GENERATED>-dep_id" refer to the one genereted text/image/audio in the dependency task (Please consider whether the dependency task generates resources of this type.) and "dep_id" must be in "dep" list. The "dep" field denotes the ids of the previous prerequisite tasks which generate a new resource that the current task relies on. The "args" field must in ["text", "image", "audio"], nothing else. The task MUST be selected from the following options: "token-classification", "text2text-generation", "summarization", "translation", "question-answering", "conversational", "text-generation", "sentence-similarity", "tabular-classification", "object-detection", "image-classification", "image-to-image", "image-to-text", "text-to-image", "text-to-video", "visual-question-answering", "document-question-answering", "image-segmentation", "depth-estimation", "text-to-speech", "automatic-speech-recognition", "audio-to-audio", "audio-classification", "canny-control", "hed-control", "mlsd-control", "normal-control", "openpose-control", "canny-text-to-image", "depth-text-to-image", "hed-text-to-image", "mlsd-text-to-image", "normal-text-to-image", "openpose-text-to-image", "seg-text-to-image". There may be multiple tasks of the same type. Think step by step about all the tasks needed to resolve the user's request. Parse out as few tasks as possible while ensuring that the user request can be resolved. Pay attention to the dependencies and order among tasks. If the user input can't be parsed, you need to reply empty JSON [].
21
  choose_model: >-
22
- #2 Model Selection Stage: Given the user request and the parsed tasks, the AI assistant helps the user to select a suitable model from a list of models to process the user request. The assistant should focus more on the description of the model and find the model that has the most potential to solve requests and tasks. Also, prefer models with local inference endpoints for speed and stability.
23
  response_results: >-
24
- #4 Response Generation Stage: With the task execution logs, the AI assistant needs to describe the process and inference results.
25
  demos_or_presteps:
26
  parse_task: demos/demo_parse_task.json
27
  choose_model: demos/demo_choose_model.json
28
  response_results: demos/demo_response_results.json
29
  prompt:
30
- parse_task: The chat log [ {{context}} ] may contain the resources I mentioned. Now I input { {{input}} }. Pay attention to the input and output types of tasks and the dependencies between tasks.
31
  choose_model: >-
32
  Please choose the most suitable model from {{metas}} for the task {{task}}. The output must be in a strict JSON format: {"id": "id", "reason": "your detail reasons for the choice"}.
33
  response_results: >-
34
- Yes. Please first think carefully and directly answer my request based on the inference results. Some of the inferences may not always turn out to be correct and require you to make careful consideration in making decisions. Then please detail your workflow including the used models and inference results for my request in your friendly tone. Please filter out information that is not relevant to my request. Tell me the complete path or urls of files in inference results. If there is nothing in the results, please tell me you can't make it. }
 
1
  openai:
2
+ key: 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
3
  huggingface:
4
+ token: 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
5
  dev: false
6
  debug: true
7
  log_file: logs/debug_TIMESTAMP.log
8
  model: text-davinci-003 # text-davinci-003
9
  use_completion: true
10
  inference_mode: hybrid # local, huggingface or hybrid
11
+ local_deployment: minimal # minimal, standard or full
12
  num_candidate_models: 5
13
  max_description_length: 100
14
  proxy:
 
17
  choose_model: 5
18
  tprompt:
19
  parse_task: >-
20
+ #1 Etap planowania zadań: Asystent AI może przeanalizować dane wprowadzane przez użytkownika do kilku zadań: [{"task": task, "id": task_id, "dep": dependency_task_id, "args": {"text": text or <GENERATED>-dep_id, "image": image_url or <GENERATED>-dep_id, "audio": audio_url or <GENERATED>-dep_id}}]. The special tag "<GENERATED>-dep_id" refer to the one genereted text/image/audio in the dependency task (Please consider whether the dependency task generates resources of this type.) and "dep_id" must be in "dep" list. The "dep" field denotes the ids of the previous prerequisite tasks which generate a new resource that the current task relies on. The "args" field must in ["text", "image", "audio"], nothing else. The task MUST be selected from the following options: "token-classification", "text2text-generation", "summarization", "translation", "question-answering", "conversational", "text-generation", "sentence-similarity", "tabular-classification", "object-detection", "image-classification", "image-to-image", "image-to-text", "text-to-image", "text-to-video", "visual-question-answering", "document-question-answering", "image-segmentation", "depth-estimation", "text-to-speech", "automatic-speech-recognition", "audio-to-audio", "audio-classification", "canny-control", "hed-control", "mlsd-control", "normal-control", "openpose-control", "canny-text-to-image", "depth-text-to-image", "hed-text-to-image", "mlsd-text-to-image", "normal-text-to-image", "openpose-text-to-image", "seg-text-to-image". Może istnieć wiele zadań tego samego typu. Pomyśl krok po kroku o wszystkich zadaniach potrzebnych do rozwiązania żądania użytkownika. Przeanalizuj jak najmniej zadań, upewniając się, że żądanie użytkownika może zostać rozwiązane. Zwróć uwagę na zależności i kolejność zadań. Jeśli dane wprowadzone przez użytkownika nie mogą zostać przeanalizowane, musisz odpowiedzieć pustym JSON [].
21
  choose_model: >-
22
+ #2 Etap wyboru modelu: Biorąc pod uwagę żądanie użytkownika i przeanalizowane zadania, asystent AI pomaga użytkownikowi wybrać odpowiedni model z listy modeli do przetworzenia żądania użytkownika. Asystent powinien bardziej skupić się na opisie modelu i znaleźć model, który ma największy potencjał do rozwiązywania próśb i zadań. Preferuj także modele z lokalnymi punktami końcowymi wnioskowania, aby uzyskać szybkość i stabilność.
23
  response_results: >-
24
+ #4 Etap generowania odpowiedzi: Za pomocą dzienników wykonania zadania asystent AI musi opisać proces i wnioskować o wynikach.
25
  demos_or_presteps:
26
  parse_task: demos/demo_parse_task.json
27
  choose_model: demos/demo_choose_model.json
28
  response_results: demos/demo_response_results.json
29
  prompt:
30
+ parse_task: The chat log [ {{context}} ] może zawierać zasoby, o których wspomniałem. Teraz wpisuję {{{input}} }. Zwróć uwagę na typy danych wejściowych i wyjściowych zadań oraz zależności między zadaniami.
31
  choose_model: >-
32
  Please choose the most suitable model from {{metas}} for the task {{task}}. The output must be in a strict JSON format: {"id": "id", "reason": "your detail reasons for the choice"}.
33
  response_results: >-
34
+ Tak. Proszę najpierw dokładnie przemyśleć i bezpośrednio odpowiedzieć na moją prośbę na podstawie wyników wnioskowania. Niektóre wnioski mogą nie zawsze okazać się poprawne i wymagać starannego rozważenia przy podejmowaniu decyzji. Następnie uprzejmie proszę szczegółowo opisać swój przepływ pracy, w tym użyte modele i wyniki wnioskowania dla mojej prośby. Proszę odfiltrować informacje, które nie istotne dla mojej prośby. Podaj pełną ścieżkę lub adresy URL plików w wynikach wnioskowania. Jeśli nie ma nic w wynikach, powiedz mi, że nie możesz tego zrobić. }