Upgrade TGI Gaudi version to v2.0.6 (#1088)

Signed-off-by: lvliang-intel <[email protected]>
Co-authored-by: chen, suyue <[email protected]>

Author: lvliang-intel

AgentQnA/docker_compose/intel/hpu/gaudi/tgi_gaudi.yaml

@@ -3,7 +3,7 @@
 
 services:
   tgi-server:
-    image: ghcr.io/huggingface/tgi-gaudi:2.0.5
+    image: ghcr.io/huggingface/tgi-gaudi:2.0.6
     container_name: tgi-server
     ports:
       - "8085:80"

AudioQnA/docker_compose/intel/hpu/gaudi/compose.yaml

@@ -51,7 +51,7 @@ services:
     environment:
       TTS_ENDPOINT: ${TTS_ENDPOINT}
   tgi-service:
-    image: ghcr.io/huggingface/tgi-gaudi:2.0.5
+    image: ghcr.io/huggingface/tgi-gaudi:2.0.6
     container_name: tgi-gaudi-server
     ports:
       - "3006:80"

AudioQnA/kubernetes/intel/README_gmc.md

@@ -25,7 +25,7 @@ The AudioQnA uses the below prebuilt images if you choose a Xeon deployment
 Should you desire to use the Gaudi accelerator, two alternate images are used for the embedding and llm services.
 For Gaudi:
 
-- tgi-service: ghcr.io/huggingface/tgi-gaudi:2.0.5
+- tgi-service: ghcr.io/huggingface/tgi-gaudi:2.0.6
 - whisper-gaudi: opea/whisper-gaudi:latest
 - speecht5-gaudi: opea/speecht5-gaudi:latest
 

AudioQnA/kubernetes/intel/hpu/gaudi/manifest/audioqna.yaml

@@ -271,7 +271,7 @@ spec:
       - envFrom:
         - configMapRef:
             name: audio-qna-config
-        image: ghcr.io/huggingface/tgi-gaudi:2.0.5
+        image: ghcr.io/huggingface/tgi-gaudi:2.0.6
         name: llm-dependency-deploy-demo
         securityContext:
           capabilities:

AudioQnA/tests/test_compose_on_gaudi.sh

@@ -22,7 +22,7 @@ function build_docker_images() {
     service_list="audioqna whisper-gaudi asr llm-tgi speecht5-gaudi tts"
     docker compose -f build.yaml build ${service_list} --no-cache > ${LOG_PATH}/docker_image_build.log
 
-    docker pull ghcr.io/huggingface/tgi-gaudi:2.0.5
+    docker pull ghcr.io/huggingface/tgi-gaudi:2.0.6
     docker images && sleep 1s
 }
 

AudioQnA/tests/test_compose_on_xeon.sh

@@ -22,7 +22,7 @@ function build_docker_images() {
     service_list="audioqna whisper asr llm-tgi speecht5 tts"
     docker compose -f build.yaml build ${service_list} --no-cache > ${LOG_PATH}/docker_image_build.log
 
-    docker pull ghcr.io/huggingface/tgi-gaudi:2.0.5
+    docker pull ghcr.io/huggingface/tgi-gaudi:2.0.6
     docker images && sleep 1s
 }
 

AvatarChatbot/docker_compose/intel/hpu/gaudi/compose.yaml

@@ -54,7 +54,7 @@ services:
     environment:
       TTS_ENDPOINT: ${TTS_ENDPOINT}
   tgi-service:
-    image: ghcr.io/huggingface/tgi-gaudi:2.0.5
+    image: ghcr.io/huggingface/tgi-gaudi:2.0.6
     container_name: tgi-gaudi-server
     ports:
       - "3006:80"

AvatarChatbot/tests/test_compose_on_gaudi.sh

@@ -29,7 +29,7 @@ function build_docker_images() {
     service_list="avatarchatbot whisper-gaudi asr llm-tgi speecht5-gaudi tts wav2lip-gaudi animation"
     docker compose -f build.yaml build ${service_list} --no-cache > ${LOG_PATH}/docker_image_build.log
 
-    docker pull ghcr.io/huggingface/tgi-gaudi:2.0.5
+    docker pull ghcr.io/huggingface/tgi-gaudi:2.0.6
 
     docker images && sleep 1s
 }

AvatarChatbot/tests/test_compose_on_xeon.sh

@@ -29,7 +29,7 @@ function build_docker_images() {
     service_list="avatarchatbot whisper asr llm-tgi speecht5 tts wav2lip animation"
     docker compose -f build.yaml build ${service_list} --no-cache > ${LOG_PATH}/docker_image_build.log
 
-    docker pull ghcr.io/huggingface/tgi-gaudi:2.0.5
+    docker pull ghcr.io/huggingface/tgi-gaudi:2.0.6
 
     docker images && sleep 1s
 }

ChatQnA/benchmark/accuracy/README.md

@@ -48,7 +48,7 @@ To setup a LLM model, we can use [tgi-gaudi](https://github.com/huggingface/tgi-
 docker run -p {your_llm_port}:80 --runtime=habana -e HABANA_VISIBLE_DEVICES=all -e PT_HPU_ENABLE_LAZY_COLLECTIVES=true -e OMPI_MCA_btl_vader_single_copy_mechanism=none -e HF_TOKEN={your_hf_token} --cap-add=sys_nice --ipc=host ghcr.io/huggingface/tgi-gaudi:2.0.1 --model-id mistralai/Mixtral-8x7B-Instruct-v0.1 --max-input-tokens 2048 --max-total-tokens 4096 --sharded true --num-shard 2
 
 # for better performance, set `PREFILL_BATCH_BUCKET_SIZE`, `BATCH_BUCKET_SIZE`, `max-batch-total-tokens`, `max-batch-prefill-tokens`
-docker run -p {your_llm_port}:80 --runtime=habana -e HABANA_VISIBLE_DEVICES=all -e PT_HPU_ENABLE_LAZY_COLLECTIVES=true -e OMPI_MCA_btl_vader_single_copy_mechanism=none -e HF_TOKEN={your_hf_token} -e PREFILL_BATCH_BUCKET_SIZE=1 -e BATCH_BUCKET_SIZE=8 --cap-add=sys_nice --ipc=host ghcr.io/huggingface/tgi-gaudi:2.0.5 --model-id mistralai/Mixtral-8x7B-Instruct-v0.1 --max-input-tokens 2048 --max-total-tokens 4096 --sharded true --num-shard 2 --max-batch-total-tokens 65536 --max-batch-prefill-tokens 2048
+docker run -p {your_llm_port}:80 --runtime=habana -e HABANA_VISIBLE_DEVICES=all -e PT_HPU_ENABLE_LAZY_COLLECTIVES=true -e OMPI_MCA_btl_vader_single_copy_mechanism=none -e HF_TOKEN={your_hf_token} -e PREFILL_BATCH_BUCKET_SIZE=1 -e BATCH_BUCKET_SIZE=8 --cap-add=sys_nice --ipc=host ghcr.io/huggingface/tgi-gaudi:2.0.6 --model-id mistralai/Mixtral-8x7B-Instruct-v0.1 --max-input-tokens 2048 --max-total-tokens 4096 --sharded true --num-shard 2 --max-batch-total-tokens 65536 --max-batch-prefill-tokens 2048
 ```
 
 ### Prepare Dataset