Fix pdf template for latest datatype refactorings

CHANGELOG.md

@@ -7,7 +7,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 **Before you create a Pull Request, remember to update the Changelog with your changes.**
 
-## Changes Since Last Release 
+## Changes Since Last Release   
 
 #### Changed defaults / behaviours 
 

plugins/sentence_transformers/superduper_sentence_transformers/model.py

@@ -2,7 +2,6 @@
 
 from sentence_transformers import SentenceTransformer as _SentenceTransformer
 from superduper.backends.query_dataset import QueryDataset
-from superduper.base.enums import DBType
 from superduper.components.component import ensure_initialized
 from superduper.components.model import Model, Signature, _DeviceManaged
 
@@ -125,4 +124,3 @@ def _pre_create(self, db):
         """
         if self.datatype is not None:
             return
-

templates/pdf_rag/component.json

@@ -63,10 +63,10 @@
         "serve": false,
         "trainer": null,
         "deploy": false,
-        "object": "&:blob:568534984ca7688b45c9a5caaa73a5756e1a78a6",
+        "object": "&:blob:fb3f160973320c9b45cb2892875aa4f2fcd33466",
         "method": null
       },
-      "sample-pdf-rag-find": {
+      "var-table-name-find": {
         "_path": "superduper_<var:databackend>.query.parse_query",
         "documents": [],
         "query": "<var:table_name>.find()"
@@ -81,7 +81,7 @@
         "key": "file",
         "model": "?model:split_image",
         "predict_kwargs": {},
-        "select": "?sample-pdf-rag-find",
+        "select": "?var-table-name-find",
         "flatten": true
       },
       "json": {
@@ -106,10 +106,10 @@
         "serve": false,
         "trainer": null,
         "deploy": false,
-        "object": "&:blob:1087871ffeeb5420a415b53f5e137701fd991584",
+        "object": "&:blob:bf6c73746c460c85c2d1cc3b419288ce434df064",
         "method": null
       },
-      "sample-pdf-rag-select": {
+      "var-table-name-select": {
         "_path": "superduper_<var:databackend>.query.parse_query",
         "documents": [],
         "query": "<var:table_name>.select()"
@@ -124,7 +124,7 @@
         "key": "file",
         "model": "?model:chunk",
         "predict_kwargs": {},
-        "select": "?sample-pdf-rag-select",
+        "select": "?var-table-name-select",
         "flatten": true
       },
       "datatype:vector[1536]": {
@@ -137,7 +137,7 @@
         "shape": [
           1536
         ],
-        "dtype": "float32"
+        "dtype": "float64"
       },
       "model:text-embedding-ada-002": {
         "_path": "superduper_openai.model.OpenAIEmbedding",
@@ -147,7 +147,7 @@
         "build_variables": null,
         "build_template": null,
         "signature": "singleton",
-        "datatype": null,
+        "datatype": "?datatype:vector[1536]",
         "output_schema": null,
         "model_update_kwargs": {},
         "predict_kwargs": {},
@@ -165,6 +165,18 @@
         "client_kwargs": {},
         "batch_size": 100
       },
+      "datatype:vector[384]": {
+        "_path": "superduper.components.datatype.Vector",
+        "upstream": null,
+        "plugins": null,
+        "cache": true,
+        "build_variables": null,
+        "build_template": null,
+        "shape": [
+          384
+        ],
+        "dtype": "float64"
+      },
       "model:sentence-transformers-embedding": {
         "_path": "superduper_sentence_transformers.model.SentenceTransformer",
         "preferred_devices": [
@@ -179,7 +191,7 @@
         "build_variables": null,
         "build_template": null,
         "signature": "singleton",
-        "datatype": null,
+        "datatype": "?datatype:vector[384]",
         "output_schema": null,
         "model_update_kwargs": {},
         "predict_kwargs": {
@@ -194,7 +206,7 @@
         "deploy": false,
         "model": "BAAI/bge-small-en",
         "preprocess": null,
-        "postprocess": "&:blob:c190af19f02c96ce9f7c38409748ccfc98758827"
+        "postprocess": "&:blob:29de8234e88613b66ede6d1dda606b6b67c4d8c8"
       },
       "model:embedding": {
         "_path": "superduper.components.model.ModelRouter",
@@ -204,7 +216,7 @@
         "build_variables": null,
         "build_template": null,
         "signature": "singleton",
-        "datatype": "?datatype:vector[1536]",
+        "datatype": null,
         "output_schema": null,
         "model_update_kwargs": {},
         "predict_kwargs": {},
@@ -219,7 +231,7 @@
           "openai": "?model:text-embedding-ada-002",
           "sentence_transformers": "?model:sentence-transformers-embedding"
         },
-        "model": "openai"
+        "model": "<var:embedding_model>"
       },
       "outputs-chunk-?(listener:chunk.uuid)-select": {
         "_path": "superduper_<var:databackend>.query.parse_query",
@@ -243,7 +255,7 @@
       },
       "vector_index:vector-index": {
         "_path": "superduper.components.vector_index.VectorIndex",
-        "upstream": null,
+        "upstream": [],
         "plugins": null,
         "cache": true,
         "build_variables": null,
@@ -261,7 +273,7 @@
         "build_variables": null,
         "build_template": null,
         "signature": "singleton",
-        "datatype": "str",
+        "datatype": null,
         "output_schema": null,
         "model_update_kwargs": {},
         "predict_kwargs": {},
@@ -280,15 +292,73 @@
         "batch_size": 1,
         "prompt": ""
       },
+      "model:llm-anthropic": {
+        "_path": "superduper_anthropic.model.AnthropicCompletions",
+        "upstream": null,
+        "plugins": null,
+        "cache": true,
+        "build_variables": null,
+        "build_template": null,
+        "signature": "*args,**kwargs",
+        "datatype": null,
+        "output_schema": null,
+        "model_update_kwargs": {},
+        "predict_kwargs": {
+          "max_tokens": 1024,
+          "temperature": 0.8
+        },
+        "compute_kwargs": {},
+        "validation": null,
+        "metric_values": {},
+        "num_workers": 0,
+        "serve": false,
+        "trainer": null,
+        "deploy": false,
+        "model": "claude-2.1",
+        "max_batch_size": 8,
+        "client_kwargs": {},
+        "prompt": ""
+      },
+      "model:llm-vllm": {
+        "_path": "superduper_vllm.model.VllmCompletion",
+        "upstream": null,
+        "plugins": null,
+        "cache": true,
+        "build_variables": null,
+        "build_template": null,
+        "signature": "*args,**kwargs",
+        "datatype": null,
+        "output_schema": null,
+        "model_update_kwargs": {},
+        "predict_kwargs": {
+          "max_tokens": 1024,
+          "temperature": 0.8
+        },
+        "compute_kwargs": {
+          "num_gpus": 1
+        },
+        "validation": null,
+        "metric_values": {},
+        "num_workers": 0,
+        "serve": false,
+        "trainer": null,
+        "deploy": false,
+        "vllm_params": {
+          "model": "TheBloke/Mistral-7B-Instruct-v0.2-AWQ",
+          "gpu_memory_utilization": 0.7,
+          "max_model_len": 1024,
+          "quantization": "awq"
+        }
+      },
       "model:llm": {
         "_path": "superduper.components.model.ModelRouter",
         "upstream": null,
         "plugins": null,
         "cache": true,
         "build_variables": null,
         "build_template": null,
-        "signature": "singleton",
-        "datatype": "str",
+        "signature": "*args,**kwargs",
+        "datatype": null,
         "output_schema": null,
         "model_update_kwargs": {},
         "predict_kwargs": {},
@@ -300,9 +370,11 @@
         "trainer": null,
         "deploy": false,
         "models": {
-          "openai": "?model:llm-openai"
+          "openai": "?model:llm-openai",
+          "anthropic": "?model:llm-anthropic",
+          "vllm": "?model:llm-vllm"
         },
-        "model": "openai"
+        "model": "<var:llm_model>"
       },
       "plugin:plugin-utils_py": {
         "_path": "superduper.components.plugin.Plugin",
@@ -339,8 +411,11 @@
         "split_image_key": "<var:output_prefix>split_image__?(listener:split_image.uuid)"
       },
       "model:rag": {
-        "_object": "&:blob:58d51953d88d26b2a08024f471d6896b12063698",
-        "upstream": null,
+        "_object": "&:blob:cd225eb265434f4caa753cfd4c1b708d67dd4d7f",
+        "upstream": [
+          "?vector_index:vector-index",
+          "?vector_index:vector-index"
+        ],
         "plugins": null,
         "cache": true,
         "build_variables": null,
@@ -358,9 +433,9 @@
         "trainer": null,
         "deploy": false,
         "llm_model": "?model:llm",
-        "vector_index_name": "vector-index",
         "prompt_template": "<var:prompt_template>",
-        "processor": "?model:processor"
+        "processor": "?model:processor",
+        "vector_index": "?vector_index:vector-index"
       },
       "pdf-rag": {
         "_path": "superduper.components.application.Application",
@@ -429,11 +504,11 @@
   "_builds": {
     "str": {
       "_path": "superduper.components.schema.FieldType",
-      "uuid": "0ba486f949584e5c"
+      "uuid": "5f1ed2f02fc64e10"
     },
     "datatype:file": {
       "_path": "superduper.components.datatype.FileType",
-      "uuid": "121dd11d21464f2b",
+      "uuid": "cb66f7db104b402b",
       "upstream": null,
       "plugins": null,
       "cache": true,
@@ -442,7 +517,7 @@
     },
     "schema:sample_pdf_rag/schema": {
       "_path": "superduper.components.schema.Schema",
-      "uuid": "a19e6f4b12914eb0",
+      "uuid": "d6afe58fc2cd4aaf",
       "upstream": null,
       "plugins": null,
       "cache": true,
@@ -456,17 +531,17 @@
     },
     "dataset:sample_pdfs": {
       "_path": "superduper.components.dataset.RemoteData",
-      "uuid": "a49980907c504046",
+      "uuid": "874a14a7699c4b2e",
       "upstream": null,
       "plugins": null,
       "cache": true,
       "build_variables": null,
       "build_template": null,
-      "getter": "&:blob:a306e0ed6eecb5fd002ef781f9b24f8d73cc44cc"
+      "getter": "&:blob:0c27f6fbe3b4538b68f0ddf6a9d2a0f8be7d1684"
     },
     "table:sample_pdf_rag": {
       "_path": "superduper.components.table.Table",
-      "uuid": "a517dec411304504",
+      "uuid": "c802c284fb7541b0",
       "upstream": null,
       "plugins": null,
       "cache": true,