Merge pull request #4 from ritual-net/sr/feat/update_services_2.0.0

stelios-ritual · web-flow · commit 686da8e52af0 · 2024-10-02T10:27:02.000-04:00
update services to v2.0.0
diff --git a/node/anvil/1.3.0/config.json b/node/anvil/1.3.0/config.json
@@ -1,6 +1,6 @@
 {
   "config": {
-    "forward_stats": false,
+    "forward_stats": true,
     "manage_containers": true,
     "startup_wait": 1.0,
     "chain": {
diff --git a/node/other/1.3.0/config.json b/node/other/1.3.0/config.json
@@ -1,6 +1,6 @@
 {
   "config": {
-    "forward_stats": false,
+    "forward_stats": true,
     "manage_containers": true,
     "startup_wait": 1.0,
     "chain": {
diff --git a/services/css-inference/2.0.0/README.md b/services/css-inference/2.0.0/README.md
diff --git a/services/css-inference/2.0.0/recipe.json b/services/css-inference/2.0.0/recipe.json
@@ -1,7 +1,7 @@
 {
     "config": {
-        "id": "css-inference-1.0.0",
-        "image": "ritualnetwork/css_inference_service:1.0.0",
+        "id": "css-inference-2.0.0",
+        "image": "ritualnetwork/css_inference_service:2.0.0",
         "description": "Closed Source Inference Service",
         "command": "--bind=0.0.0.0:3000 --workers=${NUM_WORKERS}",
         "env": {},
diff --git a/services/hf-client-inference/2.0.0/README.md b/services/hf-client-inference/2.0.0/README.md
diff --git a/services/hf-client-inference/2.0.0/recipe.json b/services/hf-client-inference/2.0.0/recipe.json
@@ -1,7 +1,7 @@
 {
     "config": {
-        "id": "hf-client-inference-1.0.0",
-        "image": "ritualnetwork/hf_inference_client_service:1.0.0",
+        "id": "hf-client-inference-2.0.0",
+        "image": "ritualnetwork/hf_inference_client_service:2.0.0",
         "description": "HuggingFace Client Inference Service",
         "command": "--bind=0.0.0.0:3000 --workers=${NUM_WORKERS}",
         "env": {},
@@ -16,11 +16,11 @@
     },
     "inputs": [
         {
-            "id": "HF_TOKEN",
-            "path": "env.HF_TOKEN",
-            "description": "The Hugging Face API token.",
+            "id": "HF_INF_TOKEN",
+            "path": "env.HF_INF_TOKEN",
+            "description": "The Hugging Face API token. Used for increased rate limits and access to private models.",
             "type": "string",
-            "required": true
+            "required": false
         },
         {
             "id": "NUM_WORKERS",
diff --git a/services/onnx-inference/2.0.0/README.md b/services/onnx-inference/2.0.0/README.md
diff --git a/services/onnx-inference/2.0.0/recipe.json b/services/onnx-inference/2.0.0/recipe.json
@@ -1,7 +1,7 @@
 {
     "config": {
-        "id": "onnx-inference-1.0.0",
-        "image": "ritualnetwork/onnx_inference_service:1.0.0",
+        "id": "onnx-inference-2.0.0",
+        "image": "ritualnetwork/onnx_inference_service:2.0.0",
         "description": "ONNX Inference Service",
         "command": "--bind=0.0.0.0:3000 --workers=${NUM_WORKERS}",
         "env": {},
@@ -16,16 +16,16 @@
     },
     "inputs": [
         {
-            "id": "MODEL_SOURCE",
-            "path": "env.MODEL_SOURCE",
-            "description": "The source of the model to load. Used for preloading a model.",
+            "id": "ONNX_DEFAULT_MODEL_ID",
+            "path": "env.ONNX_DEFAULT_MODEL_ID",
+            "description": "The default model to preload on startup.",
             "type": "string",
             "required": false
         },
         {
-            "id": "LOAD_ARGS",
-            "path": "env.LOAD_ARGS",
-            "description": "Arguments to load model with, as a stringified JSON object. Used for preloading a model.",
+            "id": "ONNX_CACHE_DIR",
+            "path": "env.ONNX_CACHE_DIR",
+            "description": "The directory to cache ONNX models in.",
             "type": "string",
             "required": false
         },
diff --git a/services/tgi-client-inference/2.0.0/README.md b/services/tgi-client-inference/2.0.0/README.md
@@ -2,6 +2,7 @@
 
 ### Description
 
-This service enables ML inference requests via HuggingFace's [Text Generation Inference (TGI)](https://huggingface.co/docs/text-generation-inference/en/index) API, which serves Large Language Models (LLMs).
+This service connects to a specified [Text Generation Inference (TGI)](https://huggingface.co/docs/text-generation-inference/en/index) instance and serves Large Language Model (LLMs) inference requests
+via the TGI REST interface.
 
 For configuration and usage details, check out the [TGI Client Inference Service](https://infernet-services.docs.ritual.net/reference/tgi_client_inference_service) documentation.
diff --git a/services/tgi-client-inference/2.0.0/recipe.json b/services/tgi-client-inference/2.0.0/recipe.json
@@ -1,11 +1,14 @@
 {
     "config": {
-        "id": "tgi-client-inference-1.0.0",
-        "image": "ritualnetwork/tgi_client_inference_service:1.0.0",
+        "id": "tgi-client-inference-2.0.0",
+        "image": "ritualnetwork/tgi_client_inference_service:2.0.0",
         "description": "TGI Client Inference Service",
         "command": "--bind=0.0.0.0:3000 --workers=${NUM_WORKERS}",
         "env": {
-            "TGI_INF_WORKFLOW_POSITIONAL_ARGS": "[\"${TGI_SERVICE_URL}\", ${CONNECTION_TIMEOUT}, {\"Authorization\": \"Bearer ${HF_TOKEN}\"}]"
+            "TGI_INF_WORKFLOW_POSITIONAL_ARGS": [
+                "${SERVICE_URL}",
+                "${CONNECTION_TIMEOUT}"
+            ]
         },
         "external": true,
         "gpu": false,
@@ -18,16 +21,16 @@
     },
     "inputs": [
         {
-            "id": "TGI_SERVICE_URL",
-            "path": "env.TGI_INF_WORKFLOW_POSITIONAL_ARGS#TGI_SERVICE_URL",
-            "description": "The URL of the TGI service to connect to.",
+            "id": "TGI_INF_TOKEN",
+            "path": "env.TGI_INF_TOKEN",
+            "description": "The Hugging Face API token. Used for increased rate limits and access to private models.",
             "type": "string",
-            "required": true
+            "required": false
         },
         {
-            "id": "HF_TOKEN",
-            "path": "env.TGI_INF_WORKFLOW_POSITIONAL_ARGS#HF_TOKEN",
-            "description": "The Hugging Face API token.",
+            "id": "SERVICE_URL",
+            "path": "env.TGI_INF_WORKFLOW_POSITIONAL_ARGS#SERVICE_URL",
+            "description": "The URL of the TGI service to connect to.",
             "type": "string",
             "required": true
         },
diff --git a/services/torch-inference/2.0.0/README.md b/services/torch-inference/2.0.0/README.md
diff --git a/services/torch-inference/2.0.0/recipe.json b/services/torch-inference/2.0.0/recipe.json
@@ -1,7 +1,7 @@
 {
     "config": {
-        "id": "torch-inference-1.0.0",
-        "image": "ritualnetwork/torch_inference_service:1.0.0",
+        "id": "torch-inference-2.0.0",
+        "image": "ritualnetwork/torch_inference_service:2.0.0",
         "description": "PyTorch Inference Service",
         "command": "--bind=0.0.0.0:3000 --workers=${NUM_WORKERS}",
         "env": {},
@@ -16,22 +16,22 @@
     },
     "inputs": [
         {
-            "id": "MODEL_SOURCE",
-            "path": "env.MODEL_SOURCE",
-            "description": "The source of the model to load. Used for preloading a model.",
+            "id": "TORCH_DEFAULT_MODEL_ID",
+            "path": "env.TORCH_DEFAULT_MODEL_ID",
+            "description": "The default model to preload on startup.",
             "type": "string",
             "required": false
         },
         {
-            "id": "LOAD_ARGS",
-            "path": "env.LOAD_ARGS",
-            "description": "Arguments to load model with, as a stringified JSON object. Used for preloading a model.",
+            "id": "TORCH_CACHE_DIR",
+            "path": "env.TORCH_CACHE_DIR",
+            "description": "The directory to cache Torch models in.",
             "type": "string",
             "required": false
         },
         {
-            "id": "USE_JIT",
-            "path": "env.USE_JIT",
+            "id": "TORCH_USE_JIT",
+            "path": "env.TORCH_USE_JIT",
             "description": "Whether to use JIT for model loading.",
             "type": "string",
             "required": false,

Original file line number	Diff line number	Diff line change
`@@ -1,6 +1,6 @@`
`1`	`1`	`{`
`2`	`2`	`"config": {`
`3`		`- "forward_stats": false,`
	`3`	`+ "forward_stats": true,`
`4`	`4`	`"manage_containers": true,`
`5`	`5`	`"startup_wait": 1.0,`
`6`	`6`	`"chain": {`