From 169d8d21ff004e31e5bf4180a06fd294a3f806e7 Mon Sep 17 00:00:00 2001
From: fakezeta <fakezeta@gmail.com>
Date: Mon, 6 May 2024 10:52:05 +0200
Subject: [PATCH] gallery: Added some OpenVINO models (#2249)

* Added some OpenVINO models

Added Phi-3 trust_remote_code: true
Added Hermes 2 Pro Llama3
Added Multilingual-E5-base embedding model with OpenVINO acceleration (CPU and XPU)
Added all-MiniLM-L6-v2 with OpenVINO acceleration (CPU and XPU)

* Added Remote Code for phi, fixed error on Yamllint

* update openvino.yaml

I need to go to rest: today is not my day...
---
 .../docs/reference/compatibility-table.md     |  5 +-
 gallery/index.yaml                            | 58 +++++++++++++++++++
 gallery/openvino.yaml                         |  3 -
 3 files changed, 61 insertions(+), 5 deletions(-)

diff --git a/docs/content/docs/reference/compatibility-table.md b/docs/content/docs/reference/compatibility-table.md
index 7c3fe154..31b47bb7 100644
--- a/docs/content/docs/reference/compatibility-table.md
+++ b/docs/content/docs/reference/compatibility-table.md
@@ -45,10 +45,11 @@ LocalAI will attempt to automatically load models which are not explicitly confi
 | [tinydream](https://github.com/symisc/tiny-dream#tiny-dreaman-embedded-header-only-stable-diffusion-inference-c-librarypixlabiotiny-dream)         | stablediffusion               | no                       | Image                 | no                                | no                   | N/A |
 | `coqui` | Coqui    | no                       | Audio generation and Voice cloning    | no                               | no                   | CPU/CUDA |
 | `petals` | Various GPTs and quantization formats | yes                      | GPT             | no | no                  | CPU/CUDA |
-| `transformers` | Various GPTs and quantization formats | yes                      | GPT, embeddings            | yes | no                  | CPU/CUDA |
+| `transformers` | Various GPTs and quantization formats | yes                      | GPT, embeddings            | yes | yes****                  | CPU/CUDA/XPU |
 
 Note: any backend name listed above can be used in the `backend` field of the model configuration file (See [the advanced section]({{%relref "docs/advanced" %}})).
 
 - \* 7b ONLY
 - ** doesn't seem to be accurate
-- *** 7b and 40b with the `ggccv` format, for instance: https://huggingface.co/TheBloke/WizardLM-Uncensored-Falcon-40B-GGML
\ No newline at end of file
+- *** 7b and 40b with the `ggccv` format, for instance: https://huggingface.co/TheBloke/WizardLM-Uncensored-Falcon-40B-GGML
+- **** Only for CUDA and OpenVINO CPU/XPU acceleration.
\ No newline at end of file
diff --git a/gallery/index.yaml b/gallery/index.yaml
index e4532547..42e62a88 100644
--- a/gallery/index.yaml
+++ b/gallery/index.yaml
@@ -1056,11 +1056,19 @@
   urls:
     - https://huggingface.co/fakezeta/Phi-3-mini-128k-instruct-ov-int8
   overrides:
+    trust_remote_code: true
     context_size: 131072
     parameters:
       model: fakezeta/Phi-3-mini-128k-instruct-ov-int8
     stopwords:
       - <|end|>
+  tags:
+    - llm
+    - openvino
+    - gpu
+    - phi3
+    - cpu
+    - Remote Code Enabled
 - <<: *openvino
   name: "openvino-starling-lm-7b-beta-openvino-int8"
   urls:
@@ -1069,6 +1077,12 @@
     context_size: 8192
     parameters:
       model: fakezeta/Starling-LM-7B-beta-openvino-int8
+  tags:
+    - llm
+    - openvino
+    - gpu
+    - mistral
+    - cpu
 - <<: *openvino
   name: "openvino-wizardlm2"
   urls:
@@ -1077,6 +1091,50 @@
     context_size: 8192
     parameters:
       model: fakezeta/Not-WizardLM-2-7B-ov-int8
+- <<: *openvino
+  name: "openvino-hermes2pro-llama3"
+  urls:
+    - https://huggingface.co/fakezeta/Hermes-2-Pro-Llama-3-8B-ov-int8
+  overrides:
+    context_size: 8192
+    parameters:
+      model: fakezeta/Hermes-2-Pro-Llama-3-8B-ov-int8
+  tags:
+    - llm
+    - openvino
+    - gpu
+    - llama3
+    - cpu
+- <<: *openvino
+  name: "openvino-multilingual-e5-base"
+  urls:
+    - https://huggingface.co/intfloat/multilingual-e5-base
+  overrides:
+    embeddings: true
+    type: OVModelForFeatureExtraction
+    parameters:
+      model: intfloat/multilingual-e5-base
+  tags:
+    - llm
+    - openvino
+    - gpu
+    - embedding
+    - cpu
+- <<: *openvino
+  name: "openvino-all-MiniLM-L6-v2"
+  urls:
+    - https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2
+  overrides:
+    embeddings: true
+    type: OVModelForFeatureExtraction
+    parameters:
+      model: sentence-transformers/all-MiniLM-L6-v2
+  tags:
+    - llm
+    - openvino
+    - gpu
+    - embedding
+    - cpu
 ### START Embeddings
 - &sentencentransformers
   description: |
diff --git a/gallery/openvino.yaml b/gallery/openvino.yaml
index b30b2a85..43c4d351 100644
--- a/gallery/openvino.yaml
+++ b/gallery/openvino.yaml
@@ -7,6 +7,3 @@ config_file: |
   type: OVModelForCausalLM
   template:
     use_tokenizer_template: true
-  stopwords:
-  - "<|eot_id|>"
-  - "<|end_of_text|>"