lex-llm-vllm 0.2.10 → 0.2.11
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +4 -4
- data/CHANGELOG.md +8 -0
- data/lib/legion/extensions/llm/vllm/provider.rb +12 -14
- data/lib/legion/extensions/llm/vllm/version.rb +1 -1
- metadata +1 -1
checksums.yaml
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
---
|
|
2
2
|
SHA256:
|
|
3
|
-
metadata.gz:
|
|
4
|
-
data.tar.gz:
|
|
3
|
+
metadata.gz: c4369ada04bb372dd59d3a57e2490c2b984c6ec1d7c4277c044b0f4126ba6136
|
|
4
|
+
data.tar.gz: 922947d49abecdbefbc7818dff56a82a35d1cae0978ebb94f5e9711f02a402c1
|
|
5
5
|
SHA512:
|
|
6
|
-
metadata.gz:
|
|
7
|
-
data.tar.gz:
|
|
6
|
+
metadata.gz: e2d9c3ffc63f2ba151573ebd898c0a08e8f73c52fe11ba38d452470e61569a33914d253ad8b1d08b235f90cfbc1f9613dea4bae5728d4b905539241e7034f9ff
|
|
7
|
+
data.tar.gz: 9cf36664ead33936f9c70ca3e892603ed6a66e1e3ef82560e7016e94207835ae6de4af7e7e9041e70eeb69b3117a48d780ca4a7e71aa3626ff8cc497800e8453
|
data/CHANGELOG.md
CHANGED
|
@@ -1,5 +1,13 @@
|
|
|
1
1
|
# Changelog
|
|
2
2
|
|
|
3
|
+
## 0.2.11 - 2026-05-21
|
|
4
|
+
|
|
5
|
+
- Add `default_transport`/`default_tier` class declarations, remove duplicate instance methods
|
|
6
|
+
- Add `model_allowed?` filtering in `discover_offerings`
|
|
7
|
+
- Identity headers included via base provider
|
|
8
|
+
- api_base reads from settings[:endpoint] fallback
|
|
9
|
+
|
|
10
|
+
|
|
3
11
|
## 0.2.10 - 2026-05-13
|
|
4
12
|
|
|
5
13
|
- Add `fetch_model_detail` to re-fetch `/v1/models` for `context_window` on a cache miss.
|
|
@@ -16,6 +16,8 @@ module Legion
|
|
|
16
16
|
class << self
|
|
17
17
|
def slug = 'vllm'
|
|
18
18
|
def local? = false
|
|
19
|
+
def default_transport = :http
|
|
20
|
+
def default_tier = :direct
|
|
19
21
|
def configuration_options = %i[vllm_api_base vllm_api_key]
|
|
20
22
|
def configuration_requirements = []
|
|
21
23
|
def capabilities = Capabilities
|
|
@@ -52,14 +54,14 @@ module Legion
|
|
|
52
54
|
end
|
|
53
55
|
|
|
54
56
|
def api_base
|
|
55
|
-
normalize_url(config.vllm_api_base || 'localhost:8000')
|
|
57
|
+
normalize_url(config.vllm_api_base || settings[:endpoint] || 'http://localhost:8000')
|
|
56
58
|
end
|
|
57
59
|
|
|
58
60
|
def headers
|
|
61
|
+
hdrs = identity_headers
|
|
59
62
|
token = config.vllm_api_key
|
|
60
|
-
|
|
61
|
-
|
|
62
|
-
{ 'Authorization' => "Bearer #{token}" }
|
|
63
|
+
hdrs['Authorization'] = "Bearer #{token}" unless token.nil? || token.to_s.empty?
|
|
64
|
+
hdrs
|
|
63
65
|
end
|
|
64
66
|
|
|
65
67
|
def health_url = '/health'
|
|
@@ -95,9 +97,13 @@ module Legion
|
|
|
95
97
|
else
|
|
96
98
|
Array(@cached_models)
|
|
97
99
|
end
|
|
98
|
-
|
|
99
|
-
|
|
100
|
+
offerings = models.filter_map do |model_info|
|
|
101
|
+
next unless model_allowed?(model_info.id)
|
|
102
|
+
|
|
103
|
+
offering_from_model(model_info)
|
|
100
104
|
end
|
|
105
|
+
log.debug { "built #{offerings.size} vLLM offering(s) live=#{live}" }
|
|
106
|
+
offerings
|
|
101
107
|
rescue StandardError => e
|
|
102
108
|
handle_exception(e, level: :warn, handled: true, operation: 'vllm.discover_offerings')
|
|
103
109
|
[]
|
|
@@ -167,14 +173,6 @@ module Legion
|
|
|
167
173
|
)
|
|
168
174
|
end
|
|
169
175
|
|
|
170
|
-
def offering_transport
|
|
171
|
-
config.respond_to?(:transport) ? config.transport : :http
|
|
172
|
-
end
|
|
173
|
-
|
|
174
|
-
def offering_tier
|
|
175
|
-
config.respond_to?(:tier) ? config.tier : :direct
|
|
176
|
-
end
|
|
177
|
-
|
|
178
176
|
def render_payload(messages, tools:, temperature:, model:, stream:, schema:, thinking:, tool_prefs:) # rubocop:disable Metrics/ParameterLists
|
|
179
177
|
payload = super
|
|
180
178
|
payload.delete(:reasoning_effort)
|