Merge pull request #352 from deeppavlov/dev

* fix: cards for deepy * fix: cards for deepy * fix: use gpt-3.5
deeppavlov · Mar 18, 2023 · a9bae9d · a9bae9d
2 parents aa44906 + 2535cc6
commit a9bae9d
Show file tree

Hide file tree

Showing 6 changed files with 52 additions and 30 deletions.
diff --git a/assistant_dists/deepy_assistant/cpu.yml b/assistant_dists/deepy_assistant/cpu.yml
@@ -8,7 +8,3 @@ services:
     environment:
       DEVICE: cpu
       CUDA_VISIBLE_DEVICES: ""
-  transformers-lm-bloomz7b:
-    environment:
-      DEVICE: cpu
-      CUDA_VISIBLE_DEVICES: ""
diff --git a/assistant_dists/deepy_assistant/dev.yml b/assistant_dists/deepy_assistant/dev.yml
@@ -45,13 +45,12 @@ services:
       - "./common:/src/common"
     ports:
       - 8135:8135
-  transformers-lm-bloomz7b:
+  openai-api-davinci3:
     volumes:
-      - "./services/transformers_lm:/src"
+      - "./services/openai_api_lm:/src"
       - "./common:/src/common"
-      - "~/.deeppavlov/cache:/root/.cache"
     ports:
-      - 8146:8146
+      - 8131:8131
   dff-deepy-prompted-skill:
     volumes:
       - "./skills/dff_template_prompted_skill:/src"

diff --git a/assistant_dists/deepy_assistant/docker-compose.override.yml b/assistant_dists/deepy_assistant/docker-compose.override.yml
@@ -3,8 +3,8 @@ services:
     command: sh -c 'bin/wait && python -m deeppavlov_agent.run agent.pipeline_config=assistant_dists/deepy_assistant/pipeline_conf.json'
     environment:
       WAIT_HOSTS: "sentseg:8011, no-restrictions-selector:8009, badlisted-words:8018, combined-classification:8087, 
-        spelling-preprocessing:8074, sentence-ranker:8128, prompt-selector:8135, transformers-lm-bloomz7b:8146, 
-        dff-deepy-prompted-skill:8148"
+                   spelling-preprocessing:8074, sentence-ranker:8128, prompt-selector:8135, openai-api-davinci3:8131,
+                   dff-deepy-prompted-skill:8148"
       WAIT_HOSTS_TIMEOUT: ${WAIT_TIMEOUT:-1000}
 
   sentseg:
@@ -145,26 +145,25 @@ services:
         reservations:
           memory: 3G
 
-  transformers-lm-bloomz7b:
+  openai-api-davinci3:
     env_file: [ .env ]
     build:
       args:
-        SERVICE_PORT: 8146
-        SERVICE_NAME: transformers_lm_bloomz7b
-        PRETRAINED_MODEL_NAME_OR_PATH: bigscience/bloomz-7b1
-        HALF_PRECISION: 0
+        SERVICE_PORT: 8131
+        SERVICE_NAME: openai_api_davinci3
+        PRETRAINED_MODEL_NAME_OR_PATH: text-davinci-003
       context: .
-      dockerfile: ./services/transformers_lm/Dockerfile
-    command: flask run -h 0.0.0.0 -p 8146
+      dockerfile: ./services/openai_api_lm/Dockerfile
+    command: flask run -h 0.0.0.0 -p 8131
     environment:
       - CUDA_VISIBLE_DEVICES=0
       - FLASK_APP=server
     deploy:
       resources:
         limits:
-          memory: 50G
+          memory: 100M
         reservations:
-          memory: 50G
+          memory: 100M
 
   dff-deepy-prompted-skill:
     env_file: [ .env ]
@@ -173,10 +172,11 @@ services:
         SERVICE_PORT: 8148
         SERVICE_NAME: dff_deepy_prompted_skill
         PROMPT_FILE: common/prompts/deepy.json
-        GENERATIVE_SERVICE_URL: http://transformers-lm-bloomz7b:8146/respond
-        GENERATIVE_SERVICE_CONFIG: default_generative_config.json
+        GENERATIVE_SERVICE_URL: http://openai-api-davinci3:8131/respond
+        GENERATIVE_SERVICE_CONFIG: openai-text-davinci-003.json
         GENERATIVE_TIMEOUT: 5
         N_UTTERANCES_CONTEXT: 3
+        ENVVARS_TO_SEND: OPENAI_API_KEY,OPENAI_ORGANIZATION
       context: .
       dockerfile: ./skills/dff_template_prompted_skill/Dockerfile
     command: gunicorn --workers=1 server:app -b 0.0.0.0:8148 --reload

diff --git a/assistant_dists/deepy_assistant/proxy.yml b/assistant_dists/deepy_assistant/proxy.yml
@@ -45,13 +45,4 @@ services:
       - PROXY_PASS=dream.deeppavlov.ai:8128
       - PORT=8128
 
-  transformers-lm-bloomz7b:
-    command: [ "nginx", "-g", "daemon off;" ]
-    build:
-      context: dp/proxy/
-      dockerfile: Dockerfile
-    environment:
-      - PROXY_PASS=dream.deeppavlov.ai:8146
-      - PORT=8146
-
 version: '3.7'
diff --git a/skills/dff_template_prompted_skill/component.yml b/skills/dff_template_prompted_skill/component.yml
@@ -206,3 +206,28 @@ dff-rhodes-coaching-prompted-skill:
     GENERATIVE_TIMEOUT: 5
     N_UTTERANCES_CONTEXT: 3
   date_created: '2023-03-16T09:45:32'
+dff-deepy-prompted-skill:
+  name: dff_deepy_prompted_skill
+  display_name: Deepy Assistant Skill
+  container_name: dff-deepy-prompted-skill
+  component_type: Script-based with NNs
+  model_type: NN-based
+  is_customizable: true
+  author: DeepPavlov
+  description: Deepy Assistant can help a user to create a customized dialogue system.
+  ram_usage: 150M
+  gpu_usage: null
+  port: 8148
+  endpoints:
+  - group: skills
+    endpoint: respond
+  build_args:
+    SERVICE_PORT: 8148
+    SERVICE_NAME: dff_deepy_prompted_skill
+    PROMPT_FILE: common/prompts/deepy.json
+    GENERATIVE_SERVICE_URL: http://openai-api-davinci3:8131/respond
+    GENERATIVE_SERVICE_CONFIG: openai-text-davinci-003.json
+    GENERATIVE_TIMEOUT: 5
+    N_UTTERANCES_CONTEXT: 3
+    ENVVARS_TO_SEND: OPENAI_API_KEY,OPENAI_ORGANIZATION
+  date_created: '2023-03-16T09:45:32'
diff --git a/skills/dff_template_prompted_skill/pipeline.yml b/skills/dff_template_prompted_skill/pipeline.yml
@@ -86,3 +86,14 @@ dff-dream-persona-davinci3-prompted-skill:
   previous_services:
   - skill_selectors
   state_manager_method: add_hypothesis
+dff-deepy-prompted-skill:
+- group: skills
+  connector:
+    protocol: http
+    timeout: 5.0
+    url: http://dff-deepy-prompted-skill:8148/respond
+  dialog_formatter: state_formatters.dp_formatters:dff_deepy_prompted_skill_formatter
+  response_formatter: state_formatters.dp_formatters:skill_with_attributes_formatter_service
+  previous_services:
+  - skill_selectors
+  state_manager_method: add_hypothesis