-
Notifications
You must be signed in to change notification settings - Fork 16
Adds TitanRefModel in place of HF based Reference Model #94
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Changes from all commits
c3f2c8c
4ca0685
3b7ee6d
41bdd93
7621fe4
29e74aa
135deaf
ff7c120
a33030d
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -13,6 +13,12 @@ | |
from typing import Dict, List | ||
|
||
import torch | ||
|
||
from forge.controller import ForgeActor, get_proc_mesh, stop_proc_mesh | ||
|
||
from forge.data.sharding import VLLMSharding | ||
from forge.interfaces import Policy as PolicyInterface | ||
from forge.types import ProcessConfig | ||
from monarch.actor import current_rank, endpoint, ProcMesh | ||
from torchstore import MultiProcessStore | ||
from torchstore._state_dict_utils import DELIM | ||
|
@@ -37,12 +43,6 @@ | |
from vllm.v1.structured_output import StructuredOutputManager | ||
from vllm.worker.worker_base import WorkerWrapperBase | ||
|
||
from forge.controller import ForgeActor, get_proc_mesh, stop_proc_mesh | ||
|
||
from forge.data.sharding import VLLMSharding | ||
from forge.interfaces import Policy as PolicyInterface | ||
from forge.types import ProcessConfig | ||
|
||
|
||
logger = logging.getLogger(__name__) | ||
|
||
|
@@ -310,7 +310,7 @@ async def run(self): | |
for request_output in processed_outputs.request_outputs: | ||
if request_output.finished: | ||
_, fut = self.requests.pop(request_output.request_id) | ||
fut.set_result(request_output.outputs) | ||
fut.set_result(request_output) | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Adopted from #97 There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Why this instead of raw outputs? There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Pragmatically: Less merge conflict with Philip's PR I don't have strong preference, but it does make the output self contained which is nice when we need to pass the results around |
||
|
||
@endpoint | ||
async def update_weights(self): | ||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
🙃