youngkingdom/vllm
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[Core][Model] Terratorch backend integration (#23513)

Browse Source 
Signed-off-by: Michele Gazzetti <michele.gazzetti1@ibm.com>
Signed-off-by: Christian Pinto <christian.pinto@ibm.com>
Co-authored-by: Christian Pinto <christian.pinto@ibm.com>
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>
This commit is contained in:
mgazz
2025-09-04 08:22:41 +01:00
committed by GitHub
parent e7fc70016f
commit 51d5e9be7d
23 changed files with 305 additions and 208 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
tests/entrypoints/openai/test_chat_template.py
View File

@ -104,7 +104,9 @@ def test_get_gen_prompt(model, template, add_generation_prompt,
trust_remote_code=model_info.trust_remote_code,
revision=model_info.revision,
hf_overrides=model_info.hf_overrides,
)
skip_tokenizer_init=model_info.skip_tokenizer_init,
enforce_eager=model_info.enforce_eager,
dtype=model_info.dtype)
# Initialize the tokenizer
tokenizer = get_tokenizer(

6
tests/entrypoints/openai/test_skip_tokenizer.py
View File

@ -11,7 +11,7 @@ import torch
from ...utils import RemoteOpenAIServer
MODEL_NAME = "christian-pinto/Prithvi-EO-2.0-300M-TL-VLLM"
MODEL_NAME = "mgazz/Prithvi-EO-2.0-300M-TL-Sen1Floods11"
DTYPE = "float16"
@ -35,7 +35,9 @@ def server():
"--trust-remote-code",
"--skip-tokenizer-init",
"--max-num-seqs",
"32"
"32",
"--model-impl",
"terratorch"
]
with RemoteOpenAIServer(MODEL_NAME, args) as remote_server:

12
tests/entrypoints/test_chat_utils.py
View File

@ -1266,7 +1266,9 @@ def test_resolve_hf_chat_template(sample_json_schema, model, use_tools):
revision=model_info.revision,
trust_remote_code=model_info.trust_remote_code,
hf_overrides=model_info.hf_overrides,
)
skip_tokenizer_init=model_info.skip_tokenizer_init,
enforce_eager=model_info.enforce_eager,
dtype=model_info.dtype)
# Build the tokenizer group and grab the underlying tokenizer
tokenizer_group = TokenizerGroup(
@ -1322,7 +1324,9 @@ def test_resolve_content_format_hf_defined(model, expected_format):
revision=model_info.revision,
trust_remote_code=model_info.trust_remote_code,
hf_overrides=model_info.hf_overrides,
)
skip_tokenizer_init=model_info.skip_tokenizer_init,
enforce_eager=model_info.enforce_eager,
dtype=model_info.dtype)
tokenizer_group = TokenizerGroup(
model,
@ -1382,7 +1386,9 @@ def test_resolve_content_format_fallbacks(model, expected_format):
revision=model_info.revision,
trust_remote_code=model_info.trust_remote_code,
hf_overrides=model_info.hf_overrides,
)
skip_tokenizer_init=model_info.skip_tokenizer_init,
enforce_eager=model_info.enforce_eager,
dtype=model_info.dtype)
tokenizer_group = TokenizerGroup(
model_config.tokenizer,