|
--- |
|
tags: |
|
- sentence-transformers |
|
- sentence-similarity |
|
- feature-extraction |
|
- generated_from_trainer |
|
- dataset_size:8000 |
|
- loss:MultipleNegativesRankingLoss |
|
base_model: sentence-transformers/all-distilroberta-v1 |
|
widget: |
|
- source_sentence: Does inhibition of autophagy by 3-MA enhance the effect of 5-FU-induced |
|
apoptosis in colon cancer cells? |
|
sentences: |
|
- Insulin- and insulin growth factor-1 stimulated signaling through the insulin |
|
receptor substrate-1 (IRS-1) promotes hepatocellular proliferation and survival. |
|
IRS-1 over-expression in transgenic (Tg) mouse livers caused constitutive activation |
|
of Erk mitogen activated protein kinase (MAPK) and phosphatidylinositol-3 kinase |
|
(PI3K) resulting in significantly increased levels of DNA synthesis and larger |
|
hepatic masses relative to non-transgenic (non-Tg) littermates. However, the livers |
|
eventually ceased to grow but remained approximately 25% larger than non-Tg livers. |
|
We hypothesized that this growth homeostasis was achieved by parallel activation |
|
of pro-apoptosis pathways. Since Fas-mediated apoptosis is a common mechanism |
|
of hepatocyte destruction, we investigated the potential role of Fas receptor |
|
as a regulator of hepatic mass in IRS-1 transgenic mice. Significantly increased |
|
Fas-receptor levels were detected in the livers of IRS-1 Tg compared to non-Tg |
|
mice by Western blot analysis. Functional activation of Fas-receptor in IRS-1 |
|
Tg livers was demonstrated by increased hepatocellular apoptosis caused by intravenous |
|
injection of anti-Fas (Jo-2). |
|
- 5-fluorouracil-(5-FU)-based adjuvant chemotherapy is widely used for the treatment |
|
of colorectal cancer. However, 5-FU resistance in the course of treatment has |
|
become more common. Therefore, new therapeutic strategies and/or new adjuvant |
|
drugs still need to be explored. Two colon-cancer-derived cell lines, colon26 |
|
and HT29, were used to investigate the effect of 5-FU, 3-methyladenine (3-MA, |
|
an autophagy inhibitor), or their combination on apoptotic cell death and autophagy. |
|
MTT assay, Hochest plus propidium iodide (PI) staining, and DNA fragmentation |
|
assay were used to observe apoptosis. Meanwhile, monodansylcadaverine (MDC) was |
|
used to detect autophagy. Finally, immunoblotting assay was used to explore the |
|
molecular change that occurred. We observed the apoptosis induced by 5-FU in colon |
|
cancer cells. Meanwhile, autophagy was also stimulated. The combination treatment |
|
of 3-MA and 5-FU significantly increased the apoptotic cell death. By isolating |
|
the subcellular fractions of mitochondria and cytosol, we observed that the release |
|
of cytochrome c was increased in combination-treated cells. Cytochrome c resulted |
|
in the activation of caspase-3, thus activating PARP. Moreover, the anti-apoptotic |
|
protein, Bcl-xL, was significantly downregulated by 3-MA. |
|
- To measure the serum highly sensitive C-reactive protein (hs-CRP) and adiponectin |
|
levels, assess insulin sensitivity index (SI) and acute insulin response (AIR) |
|
in normal control (NC) subjects, patients with impaired glucose tolerance (IGT) |
|
and newly diagnosed type 2 diabetes mellitus (DM), and further explore the possible |
|
correlation between hs-CRP and SI, AIR and adiponectin in IGT and newly diagnosed |
|
type 2 DM groups. Age and sex matched 28 normal subjects, 31 patients with IGT, |
|
and 31 patients with newly diagnosed type 2 DM were included in the study. SI |
|
and AIR were assessed by the reduced sample number of Bergman's minimal model |
|
method with intravenous glucose tolerance test in subjects of each group. Compared |
|
with NC group, serum hs-CRP was significantly increased in IGT and type 2 DM groups |
|
(p < 0.001), although there was no significant difference between the latter groups. |
|
Hs-CRP was negatively correlated with high density lipoprotein cholesterol (HDL-C), |
|
SI and adiponectin levels (p < 0.05 to p < 0.001), and positively correlated with |
|
systolic blood pressure (SBP), fasting plasma glucose (FPG), BMI, waist-to-hip |
|
ratio (WHR), postprandial 2h plasma glucose (2hPG), fasting serum insulin (FINS) |
|
and postprandial serum insulin (PSI) in IGT and newly diagnosed type 2 DM groups |
|
(p < 0.05 to p < 0.001). In general multivariate regression, only adiponectin |
|
was the significantly independent determinant for serum hs-CRP (regression coefficient |
|
-1.380; 95% CI -2.062 to 0.698, p < 0.001); meanwhile, TG, SI, hs-CRP, FINS, 2hPG |
|
and WHR were significantly independent determinants for serum adiponectin concentration |
|
(p < 0.05 to p < 0.001). |
|
- source_sentence: Is myostatin expression altered by insulin deficiency and replacement |
|
in streptozotocin-diabetic rat skeletal muscles? |
|
sentences: |
|
- 'Insulin is a major post-prandial muscle-anabolic hormone. A substantial loss |
|
of skeletal muscle mass occurs in insulin-deprived diabetes and is reversed by |
|
insulin treatment. Myostatin is a negative regulator of muscle mass upregulated |
|
in several chronic catabolic conditions. Whether myostatin expression is altered |
|
in insulin-deprived diabetes is unknown. In spite of opposite effects on muscle |
|
mass the potential role of basal circulating insulin in the regulation of myostatin |
|
expression is also undetermined. We measured (Northern Blot) myostatin transcript |
|
levels in muscle groups with different fiber composition in streptozotocin-diabetic |
|
male rats receiving one of the following treatments for eight weeks: (1) control |
|
(C); (2) diabetes without treatment (DM); (3) diabetes with once-daily slow-acting |
|
insulin treatment (INS). INS normalized plasma insulin and prevented weight reduction |
|
observed in DM. In fast-twitch gastrocnemius muscle myostatin transcript levels |
|
were unchanged (P>0.4) in both DM and INS compared to C. Myostatin transcripts |
|
were not measurable in any group in slow-twitch soleus muscle.' |
|
- A 'cannot intubate-cannot ventilate' situation requires emergency insertion of |
|
an infraglottic surgical airway. We present a case of postoperative macroglossia |
|
requiring emergency insertion of an uncuffed percutaneous cricothyroidotomy tube. |
|
The supraglottic leak was eliminated by the insertion of a laryngeal mask airway |
|
with an occluded 15-mm connector. A 49-yr-old man underwent clipping of a left |
|
posterior inferior cerebellar artery aneurysm and his tracheal tube was removed |
|
postoperatively. Two hours later, he became dyspneic and developed significant |
|
macroglossia. After application of topical anesthesia, direct laryngoscopy, oral |
|
fibreoptic bronchoscopy and laryngeal mask insertion were unsuccessful. The patient |
|
became progressively hypoxemic, pulseless electrical activity ensued, and cardiopulmonary |
|
resuscitation was initiated. An uncuffed percutaneous cricothyroidotomy tube was |
|
inserted. Oxygenation and hemodynamics were restored. As the cricothyroidotomy |
|
tube was uncuffed, there was a large supraglottic leak with manual ventilation. |
|
A laryngeal mask airway was inserted and the cuff was inflated. The 15-mm connector |
|
was occluded by a piece of tape. Subsequently, there was no further supraglottic |
|
leak with manual ventilation. He was taken to operating room and a surgical tracheotomy |
|
was performed. |
|
- 'Sepsis is an emergency condition with high mortality and morbidity rate. There |
|
are limited data on the association of cancer as a risk factor for mortality in |
|
sepsis patients in the emergency department (ED). This retrospective study was |
|
conducted at the ED, Faculty of Medicine, Ramathibodi Hospital, Bangkok, Thailand. |
|
The study period was between January 1st and December 31st, 2014. The inclusion |
|
criteria were as follows: adult patients over 15 years of age who presented at |
|
the ED with suspicion of sepsis, received treatment at the ED, and whose blood |
|
culture was found to be positive. Clinical data were recorded from medical records |
|
including the Mortality in Emergency Department Sepsis score (MEDS score). The |
|
primary outcome of this study was mortality at one month. Multivariate logistic |
|
regression analysis was used to identify independent factors associated with death. |
|
During the study period, there were 775 eligible patients. The two most common |
|
pathogens identified from blood cultures were Staphylococcus aureus (193 patients; |
|
24.9%) and Escherichia coli (158 patients; 20.4%). At one month after presenting |
|
at the ED, 110 patients (14.2%) had died. There were four significant factors |
|
for death, having cancer, being on an endotracheal tube, initial diagnosis of |
|
bacteremia, and high MED scores. Having cancer had an adjusted OR of 2.12 (95% |
|
CI of 1.29, 3.47).' |
|
- source_sentence: Is elevated plasma aldosterone an independent risk factor for erectile |
|
dysfunction in men? |
|
sentences: |
|
- The aims of this study were to demonstrate the tumorigenicity of CD133+ colon |
|
cancer cells in vitro, analyze the correlations between spheroid formation and |
|
clinicopathologic variables, and screen for overexpressed genes in CD133+ colon |
|
cancer stem cells. Moreover, the aim of this study was to establish a living tumor |
|
tissue bank using surgically resected specimens. Using LoVo cell line, we isolated |
|
CD133+ cells and performed clonogenic assay and animal experiment to test tumorigenicity |
|
of CD133+ cells. Twenty-nine surgical samples were freshly collected from 27 patients |
|
who received curative or palliative surgery, and the samples were mechanically |
|
and enzymatically dissociated into single cells. We confirmed the enhanced tumorigenicity |
|
of CD133+ cells isolated from LoVo cell line both in vitro and in vivo. Of these |
|
29 samples, 8 (28%) contained >3% CD133+ cells. Sphere formation was significantly |
|
higher in samples from patients with lymphatic invasion than in those without |
|
lymphatic invasion [54.5% (6/11) vs. 12.5% (2/16); P=0.033] and in samples containing |
|
>3% of CD133+ cells than in those containing ≤3% of CD133+ cells [36.4% (4/11) |
|
vs. 0% (0/16); P=0.019]. |
|
- Erectile dysfunction (ED) and cardiovascular disease (CVD) share a great number |
|
of common risk factors. There is growing evidence that aldosterone, an independent |
|
CVD risk factor, is associated with ED. The purpose of this study was to determine |
|
the relationship between plasma aldosterone and erectile dysfunction. This study |
|
recruited 287 participants, ranging from 18 to 84 years old; 217 were suffering |
|
from ED, diagnosed by the International Index of Erectile Function 5 (IIEF-5) |
|
scores. Based on IIEF-5 scores, patients were divided into one control group and |
|
three ED groups (mild ED; moderate ED; severe ED). The differences in principal |
|
characteristics, blood routine, sexual hormone, adrenal hormone, thyroid hormone, |
|
renal function, liver function and blood lipid were compared between ED and control |
|
groups. Our study demonstrated that the difference of mean plasma aldosterone |
|
levels between ED group and the control group was statistically significant (P < 0.05). |
|
Stepwise logistic regression analysis of all the possible factors support the |
|
role of aldosterone as an independent risk factor for ED (OR 1.011; 95 % CI 1.003-1.018; |
|
P = 0.004). Similar statistical methods were applied to the comparison between |
|
moderate to severe ED group and control to mild ED group (OR 1.017; 95 % CI 1.009-1.024; |
|
P < 0.001). ROC curve and the area under the curve (0.718; 95 % CI 0.643-0.794; |
|
P < 0.001) were performed to assess the diagnostic effect and to compare the severity |
|
of risk with the known independent risk factors, such as age and cholesterol (0.704; |
|
95 % CI 0.631-0.778; P < 0.001). When using a 374 pg/mL cut-off value from Youden |
|
index, the OR of ED group versus controls is 3.106 (95 % CI 1.458-6.617), while |
|
the OR of moderate to severe ED versus control and mild ED is 5.480 (95 % CI 3.108-9.662). |
|
- 'The objective of this study was to examine clinical outcomes and recanalization |
|
rates in a multicenter cohort of stroke patients receiving intravenous tissue |
|
plasminogen activator by site of occlusion localized with bedside transcranial |
|
Doppler. Angiographic studies with intraarterial thrombolysis suggest more proximal |
|
occlusions carry greater thrombus burden and benefit less from local therapy. |
|
Using validated transcranial Doppler criteria for specific arterial occlusion |
|
(Thrombolysis in Brain Ischemia flow grades), we compared the rate of dramatic |
|
recovery (National Institutes of Health Stroke Scale score < or =2 at 24 hours) |
|
and favorable outcomes at 3 months (modified Rankin Scale < or =1) for each occlusion |
|
site. We determined the likelihood of recanalization at various occlusion sites |
|
and its predictors. Then, stepwise logistic regression was used to determine predictors |
|
of complete recanalization. Three hundred thirty-five patients had a mean age |
|
69+/-13 years and 48.5% were women (median baseline National Institutes of Health |
|
Stroke Scale score 16 [range, 3 to 32], mean time to transcranial Doppler 140+/-84 |
|
minutes, and mean time to intravenous tissue plasminogen activator 145+/-68 minutes). |
|
Distal middle cerebral artery occlusion had an OR of 2 for complete recanalization |
|
(50 of 113 [44.2%], 95% CI: 1.1 to 3.1, P=0.005), proximal middle cerebral artery |
|
0.7 (49 of 163 [30%], 95% CI: 0.4 to 1.1, P=0.13), terminal internal carotid artery |
|
0.1 (one of 17 [5.9%], 95% CI: 0.015 to 0.8, P=0.015), tandem cervical internal |
|
carotid artery/middle cerebral artery 0.7 (6 of 22 [27%], 95% CI: 0.3 to 1.9, |
|
P=0.5), and basilar artery 0.96 (3 of 10 [30%], 95% CI: 0.2 to 4, P=0.9). Prerecombinant |
|
tissue plasminogen activator National Institutes of Health Stroke Scale score, |
|
systolic blood pressure, glucose, and Thrombolysis in Brain Ischemia flow grade |
|
at the occlusion site were the negative independent predictors for complete recanalization |
|
in the final model. There were no associations among time to treatment, stroke |
|
mechanisms, or recanalization rate. Patients with no flow (Thrombolysis in Brain |
|
Ischemia 0) at the occlusion site had less probability of complete recanalization |
|
than patients with dampened flow (Thrombolysis in Brain Ischemia 3) (OR(adj): |
|
0.256, 95% CI: 0.11 to 0.595, P=0.002). Continuous transcranial Doppler monitoring |
|
(exposure to ultrasound) was a positive predictor for complete recanalization |
|
(OR(adj): 3.02, 95% CI: 1.396 to 6.514, P=0.005). National Institutes of Health |
|
Stroke Scale score < or =2 at 24 hours was achieved in 66 of 305 patients (22%): |
|
distal middle cerebral artery 33% (35 of 107), tandem cervical internal carotid |
|
artery/middle cerebral artery 24% (5 of 21), proximal middle cerebral artery 16% |
|
(24 of 155), basilar artery 25% (2 of 8), and none of the patients with terminal |
|
internal carotid artery had dramatic recovery (0%, n=14; P=0.003). Modified Rankin |
|
Scale score < or =1 was achieved in 90 of 260 patients (35%): distal middle cerebral |
|
artery 52% (50 of 96), proximal middle cerebral artery 25% (33 of 131), tandem |
|
cervical internal carotid artery/middle cerebral artery 21% (3 of 14), terminal |
|
internal carotid artery 18% (2 of 11), and basilar artery 25% (2 of 8) (P<0.001). |
|
Patients with distal middle cerebral artery occlusion were twice as likely to |
|
have a good long-term outcome as patients with proximal middle cerebral artery |
|
(OR: 2.1, 95% CI: 1.1 to 4, P=0.025).' |
|
- source_sentence: Is the Artificial Urinary Sphincter Superior to a Secondary Transobturator |
|
Male Sling in Cases of a Primary Sling Failure? |
|
sentences: |
|
- We compared continence outcomes in patients with post-prostatectomy stress urinary |
|
incontinence treated with a salvage artificial urinary sphincter vs a secondary |
|
transobturator sling. We retrospectively reviewed the records of patients undergoing |
|
salvage procedures after sling failure from 2006 to 2012. Postoperative success |
|
was defined as the use of 0 or 1 pad, a negative stress test and pad weight less |
|
than 8 gm per day. We performed the Wilcoxon test and used a Cox regression model |
|
and Kaplan-Meier survival analysis. A total of 61 men presenting with sling failure |
|
were included in study, of whom 32 went directly to an artificial urinary sphincter |
|
and 29 received a secondary sling. Of the artificial urinary sphincter cohort |
|
47% underwent prior external beam radiation therapy vs 17% of the secondary sling |
|
cohort (p = 0.01). Average preoperative 24 hour pad weight and pad number were |
|
higher in the artificial urinary sphincter cohort. Median followup in artificial |
|
urinary sphincter and secondary sling cases was 4.5 (IQR 4-12) and 4 months (IQR |
|
1-5), respectively. Overall treatment failure was seen in 55% of patients (16 |
|
of 29) with a secondary sling vs 6% (2 of 32) with an artificial urinary sphincter |
|
(unadjusted HR 7, 95% CI 2-32 and adjusted HR 6, 95% CI 1-31). |
|
- Subcortical band heterotopia (SBH) is a cortical malformation formed when neocortical |
|
neurons prematurely stop their migration in the white matter, forming a heterotopic |
|
band below the normotopic cortex, and is generally associated with intractable |
|
epilepsy. Although it is clear that the band heterotopia and the overlying cortex |
|
both contribute to creating an abnormal circuit prone to generate epileptic discharges, |
|
it is less understood which part of this circuitry is the most critical. Here, |
|
we sought to identify the origin of epileptiform activity in a targeted genetic |
|
model of SBH in rats. Rats with SBH (Dcx-KD rats) were generated by knocking down |
|
the Dcx gene using shRNA vectors transfected into neocortical progenitors of rat |
|
embryos. Origin, spatial extent, and laminar profile of bicuculline-induced interictal-like |
|
activity on neocortical slices were analyzed by using extracellular recordings |
|
from 60-channel microelectrode arrays. Susceptibility to pentylenetetrazole-induced |
|
seizures was assessed by electrocorticography in head-restrained nonanesthetized |
|
rats. We show that the band heterotopia does not constitute a primary origin for |
|
interictal-like epileptiform activity in vitro and is dispensable for generating |
|
induced seizures in vivo. Furthermore, we report that most interictal-like discharges |
|
originating in the overlying cortex secondarily propagate to the band heterotopia. |
|
Importantly, we found that in vivo suppression of neuronal excitability in SBH |
|
does not alter the higher propensity of Dcx-KD rats to display seizures. |
|
- 'Cardiac glycosides such as digoxin and their endogenous counterpart digoxin-like |
|
immunoreactive factor (DLIF) may possess anti-inflammatory properties. Pro-inflammatory |
|
cytokines from human peripheral blood mononuclear cells (PBMC) were measured by |
|
ELISA using specific antibodies. Immunocytochemistry was used to localize NF-K: |
|
B. Non-stimulated PBMC constitutively secreted minimum amounts of cytokines. LPS |
|
(1 mg/L) stimulation lead to steep increases in TNF-alpha, IL-6 and IL-8 concentrations |
|
with peak rises at 8 h. An 8 h delay was observed for IL-10. Increases in IL-10 |
|
were sustained for18 h period. Significant inhibition (P > 0.05) of TNF-alpha, |
|
IL-6 and IL-8 at non-toxic of digoxin concentration (< 100 nM) and DLIF (10 nM |
|
digoxin equivalent (de)) was observed whereas no such effect was seen for IL-10. |
|
Inhibition of the degradation of activated NF-K: B in the PBMC was observed with |
|
the indicated concentrations of digoxin, DLIF or Pyrrolidine dithiocarbamate (PDTC).' |
|
- source_sentence: Do transport properties of pancreatic cancer describe gemcitabine |
|
delivery and response? |
|
sentences: |
|
- Hormone receptor-positive, pre-menopausal breast cancer patients can be treated |
|
by chemotherapy and/or ovarian suppression therapy. We reported our experience |
|
of gonadotropin-releasing hormone analogue plus tamoxifen (GnRHa+T) or adriamycin |
|
and cyclophosphamide (AC) followed by tamoxifen (AC-->T) in pre-menopausal women |
|
with hormone-response, node-negative breast cancer. We retrospectively reviewed |
|
the records of 587 pre-menopausal women with hormone-responsive, node-negative |
|
breast cancer. Of these, 269 were treated with adriamycin and cyclophosphamide |
|
(AC) followed by tamoxifen (AC-->T), and 318 were treated with gonadotropin-releasing |
|
hormone analogue plus tamoxifen (GnRHa+T). Among them, 151 patients were treated |
|
by goserelin acetate 3.6 mg/kg and 125 patients were treated by leuprorelin acetate |
|
3.75 mg/kg every 28 days subcutaneously. At a median follow-up time of 30 months, |
|
eight patients had relapsed and three had died. DFS did not differ between the |
|
AC-->T and GnRHa+T groups. Of the three deaths, two were not related to breast |
|
cancer. The third patient, in the AC-->T group, died because of brain metastasis. |
|
GnRHa+T treatment had no effect on blood profile and did not cause the development |
|
of detrimental symptoms but decreased bone mineral density. The efficacy of leuprorelin |
|
was similar to that of goserelin. |
|
- Atopy and systemic onset juvenile idiopathic arthritis (SoJIA) are two potential |
|
outcomes of a dysregulated immune system. Although rare, SoJIA causes 60% of the |
|
morbidity of JIA patients which exhibit a wide heterogeneity of prognosis and |
|
treatment. Co-morbidities can complicate the responses to therapy. To study the |
|
influence of co-existing atopy on the prognosis of SoJIA. Patients diagnosed with |
|
SoJIA between Jan 2006 and Sep 2010 were screened, enrolled in this prospective |
|
cohort study, and followed for 2 years. Management of SoJIA patients was assessed |
|
by ACR Pedi30/50/70 criteria, laboratory variables, and systemic feature score. |
|
At disease onset, 61 SoJIA patients (34 male and 27 female) were enrolled and |
|
were divided into SoJIA patients with atopy (n = 27) or those without atopy (n = 34). |
|
Atopic group at disease onset had significantly higher numbers of affected joints, |
|
ferritin levels and IgE serum levels than the non-atopic group. At 3 and 6 months, |
|
fewer SoJIA patients with atopy reached the ACR Pedi50 criteria (p < 0.02). During |
|
the 2 years of follow-up time, the number of infections and the number of flares |
|
were significantly higher in the SoJIA with atopy group (p < 0.01). |
|
- The therapeutic resistance of pancreatic ductal adenocarcinoma (PDAC) is partly |
|
ascribed to ineffective delivery of chemotherapy to cancer cells. We hypothesized |
|
that physical properties at vascular, extracellular, and cellular scales influence |
|
delivery of and response to gemcitabine-based therapy. We developed a method to |
|
measure mass transport properties during routine contrast-enhanced CT scans of |
|
individual human PDAC tumors. Additionally, we evaluated gemcitabine infusion |
|
during PDAC resection in 12 patients, measuring gemcitabine incorporation into |
|
tumor DNA and correlating its uptake with human equilibrative nucleoside transporter |
|
(hENT1) levels, stromal reaction, and CT-derived mass transport properties. We |
|
also studied associations between CT-derived transport properties and clinical |
|
outcomes in patients who received preoperative gemcitabine-based chemoradiotherapy |
|
for resectable PDAC. Transport modeling of 176 CT scans illustrated striking differences |
|
in transport properties between normal pancreas and tumor, with a wide array of |
|
enhancement profiles. Reflecting the interpatient differences in contrast enhancement, |
|
resected tumors exhibited dramatic differences in gemcitabine DNA incorporation, |
|
despite similar intravascular pharmacokinetics. Gemcitabine incorporation into |
|
tumor DNA was inversely related to CT-derived transport parameters and PDAC stromal |
|
score, after accounting for hENT1 levels. Moreover, stromal score directly correlated |
|
with CT-derived parameters. Among 110 patients who received preoperative gemcitabine-based |
|
chemoradiotherapy, CT-derived parameters correlated with pathological response |
|
and survival. |
|
datasets: |
|
- subhrajit-mohanty/PubMedQA_instruction_fotmatted |
|
pipeline_tag: sentence-similarity |
|
library_name: sentence-transformers |
|
metrics: |
|
- cosine_accuracy |
|
model-index: |
|
- name: SentenceTransformer based on sentence-transformers/all-distilroberta-v1 |
|
results: |
|
- task: |
|
type: triplet |
|
name: Triplet |
|
dataset: |
|
name: ai pubmed validation |
|
type: ai-pubmed-validation |
|
metrics: |
|
- type: cosine_accuracy |
|
value: 1.0 |
|
name: Cosine Accuracy |
|
--- |
|
|
|
# SentenceTransformer based on sentence-transformers/all-distilroberta-v1 |
|
|
|
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [sentence-transformers/all-distilroberta-v1](https://huggingface.co/sentence-transformers/all-distilroberta-v1) on the [pub_med_qa_instruction_fotmatted](https://huggingface.co/datasets/subhrajit-mohanty/PubMedQA_instruction_fotmatted) dataset. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more. |
|
|
|
## Model Details |
|
|
|
### Model Description |
|
- **Model Type:** Sentence Transformer |
|
- **Base model:** [sentence-transformers/all-distilroberta-v1](https://huggingface.co/sentence-transformers/all-distilroberta-v1) <!-- at revision 842eaed40bee4d61673a81c92d5689a8fed7a09f --> |
|
- **Maximum Sequence Length:** 512 tokens |
|
- **Output Dimensionality:** 768 dimensions |
|
- **Similarity Function:** Cosine Similarity |
|
- **Training Dataset:** |
|
- [pub_med_qa_instruction_fotmatted](https://huggingface.co/datasets/subhrajit-mohanty/PubMedQA_instruction_fotmatted) |
|
<!-- - **Language:** Unknown --> |
|
<!-- - **License:** Unknown --> |
|
|
|
### Model Sources |
|
|
|
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net) |
|
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers) |
|
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers) |
|
|
|
### Full Model Architecture |
|
|
|
``` |
|
SentenceTransformer( |
|
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: RobertaModel |
|
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True}) |
|
(2): Normalize() |
|
) |
|
``` |
|
|
|
## Usage |
|
|
|
### Direct Usage (Sentence Transformers) |
|
|
|
First install the Sentence Transformers library: |
|
|
|
```bash |
|
pip install -U sentence-transformers |
|
``` |
|
|
|
Then you can load this model and run inference. |
|
```python |
|
from sentence_transformers import SentenceTransformer |
|
|
|
# Download from the 🤗 Hub |
|
model = SentenceTransformer("sentence_transformers_model_id") |
|
# Run inference |
|
sentences = [ |
|
'Do transport properties of pancreatic cancer describe gemcitabine delivery and response?', |
|
'The therapeutic resistance of pancreatic ductal adenocarcinoma (PDAC) is partly ascribed to ineffective delivery of chemotherapy to cancer cells. We hypothesized that physical properties at vascular, extracellular, and cellular scales influence delivery of and response to gemcitabine-based therapy. We developed a method to measure mass transport properties during routine contrast-enhanced CT scans of individual human PDAC tumors. Additionally, we evaluated gemcitabine infusion during PDAC resection in 12 patients, measuring gemcitabine incorporation into tumor DNA and correlating its uptake with human equilibrative nucleoside transporter (hENT1) levels, stromal reaction, and CT-derived mass transport properties. We also studied associations between CT-derived transport properties and clinical outcomes in patients who received preoperative gemcitabine-based chemoradiotherapy for resectable PDAC. Transport modeling of 176 CT scans illustrated striking differences in transport properties between normal pancreas and tumor, with a wide array of enhancement profiles. Reflecting the interpatient differences in contrast enhancement, resected tumors exhibited dramatic differences in gemcitabine DNA incorporation, despite similar intravascular pharmacokinetics. Gemcitabine incorporation into tumor DNA was inversely related to CT-derived transport parameters and PDAC stromal score, after accounting for hENT1 levels. Moreover, stromal score directly correlated with CT-derived parameters. Among 110 patients who received preoperative gemcitabine-based chemoradiotherapy, CT-derived parameters correlated with pathological response and survival.', |
|
'Atopy and systemic onset juvenile idiopathic arthritis (SoJIA) are two potential outcomes of a dysregulated immune system. Although rare, SoJIA causes 60% of the morbidity of JIA patients which exhibit a wide heterogeneity of prognosis and treatment. Co-morbidities can complicate the responses to therapy. To study the influence of co-existing atopy on the prognosis of SoJIA. Patients diagnosed with SoJIA between Jan 2006 and Sep 2010 were screened, enrolled in this prospective cohort study, and followed for 2 years. Management of SoJIA patients was assessed by ACR Pedi30/50/70 criteria, laboratory variables, and systemic feature score. At disease onset, 61 SoJIA patients (34 male and 27 female) were enrolled and were divided into SoJIA patients with atopy (n\u2009=\u200927) or those without atopy (n\u2009=\u200934). Atopic group at disease onset had significantly higher numbers of affected joints, ferritin levels and IgE serum levels than the non-atopic group. At 3 and 6 months, fewer SoJIA patients with atopy reached the ACR Pedi50 criteria (p\u2009<\u20090.02). During the 2 years of follow-up time, the number of infections and the number of flares were significantly higher in the SoJIA with atopy group (p\u2009<\u20090.01).', |
|
] |
|
embeddings = model.encode(sentences) |
|
print(embeddings.shape) |
|
# [3, 768] |
|
|
|
# Get the similarity scores for the embeddings |
|
similarities = model.similarity(embeddings, embeddings) |
|
print(similarities.shape) |
|
# [3, 3] |
|
``` |
|
|
|
<!-- |
|
### Direct Usage (Transformers) |
|
|
|
<details><summary>Click to see the direct usage in Transformers</summary> |
|
|
|
</details> |
|
--> |
|
|
|
<!-- |
|
### Downstream Usage (Sentence Transformers) |
|
|
|
You can finetune this model on your own dataset. |
|
|
|
<details><summary>Click to expand</summary> |
|
|
|
</details> |
|
--> |
|
|
|
<!-- |
|
### Out-of-Scope Use |
|
|
|
*List how the model may foreseeably be misused and address what users ought not to do with the model.* |
|
--> |
|
|
|
## Evaluation |
|
|
|
### Metrics |
|
|
|
#### Triplet |
|
|
|
* Dataset: `ai-pubmed-validation` |
|
* Evaluated with [<code>TripletEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.TripletEvaluator) |
|
|
|
| Metric | Value | |
|
|:--------------------|:--------| |
|
| **cosine_accuracy** | **1.0** | |
|
|
|
<!-- |
|
## Bias, Risks and Limitations |
|
|
|
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.* |
|
--> |
|
|
|
<!-- |
|
### Recommendations |
|
|
|
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.* |
|
--> |
|
|
|
## Training Details |
|
|
|
### Training Dataset |
|
|
|
#### pub_med_qa_instruction_fotmatted |
|
|
|
* Dataset: [pub_med_qa_instruction_fotmatted](https://huggingface.co/datasets/subhrajit-mohanty/PubMedQA_instruction_fotmatted) at [66d9473](https://huggingface.co/datasets/subhrajit-mohanty/PubMedQA_instruction_fotmatted/tree/66d9473c747f2e73a050b733f8a871cf37281b93) |
|
* Size: 8,000 training samples |
|
* Columns: <code>instruction</code>, <code>context</code>, and <code>context_neg</code> |
|
* Approximate statistics based on the first 1000 samples: |
|
| | instruction | context | context_neg | |
|
|:--------|:-----------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------| |
|
| type | string | string | string | |
|
| details | <ul><li>min: 12 tokens</li><li>mean: 26.21 tokens</li><li>max: 56 tokens</li></ul> | <ul><li>min: 20 tokens</li><li>mean: 320.41 tokens</li><li>max: 512 tokens</li></ul> | <ul><li>min: 24 tokens</li><li>mean: 319.78 tokens</li><li>max: 512 tokens</li></ul> | |
|
* Samples: |
|
| instruction | context | context_neg | |
|
|:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| |
|
| <code>Do adipose-derived mesenchymal stem cells ameliorate STZ-induced pancreas damage in type 1 diabetes?</code> | <code>To investigate the possibility of adipose-derived mesenchymal stem cells (ADSC) in the treatment of type 1 diabetes (T1D). ADSC were isolated from the adipotic tissue of abdomen in Sprague-Dawley rats (4-6 week-old,female) and expanded in vitro. Cells were then identified by testing their phenotypes through flow cytometry. Balb/c mice (8 week-old, male) were divided into 3 groups: T1D group, ADSC group and control group. Streptozocin (50 mg/kg·d) were injected intraperitoneally into mice of T1D group and ADSC group for 5 consecutive days to establish the T1D model. In ADSC group, ADSC were injected intravenously on day 3 of STZ injection. In control group, only PBS was injected. Fasting blood glucose (FGB) level was examined once a week. At the end of the 4th week, animals were killed. The pathological changes of islet were showed by histochemistry through hematoxylin-eosin staining (HE staining). β cell insulin expression was detected by quantum dots immunofluorescence histochemistry....</code> | <code>We conducted this pooled analysis to assess the prognostic value of pretreatment Quality of Life (QOL) assessments on overall survival (OS) in advanced non-small cell lung cancer (NSCLC). Four hundred twenty patients with advanced NSCLC (stages IIIB with pleural effusion and IV) from six North Central Cancer Treatment Group trials were included in this study. QOL assessments included the single-item Uniscale (355 patients), Lung Cancer Symptom Scale (217 patients), and Functional Assessment of Cancer Therapy-Lung (197 patients). QOL scores were transformed to a 0 to 100 scale with higher scores representing better status and categorized using the sample median or clinically deficient score (CDS, <or=50 versus >50). Cox proportional hazards models stratified by study were used to evaluate the prognostic importance of QOL on OS alone and in the presence of other prognostic factors such as performance status, age, gender, body mass index, and laboratory parameters. Pretreatment QOL access...</code> | |
|
| <code>Is prevalence of asthma and allergic diseases in Croatian children increasing : survey study?</code> | <code>To estimate the prevalence of asthma, allergic rhinitis, and atopic dermatitis among school children in the region of Primorsko-goranska County in Croatia, and compare the results with data from other countries. The study was conducted during the 2001-2002 school year, in complete adherence to the Phase One protocol of the International Study of Asthma and Allergies in Childhood (ISAAC). The target population comprised two age groups (6-7 and 13-14 years) in the region of Primorsko-Goranska County in Croatia. Data were collected using standardized ISAAC written questionnaire and asthma video questionnaire. There were 1,634 participating children in the 6-7 age group (response rate 80.3%) and 2,194 participating children in the 13-14 age group (response rate 89.8%). Estimated 12-month prevalence rates of symptoms were: wheezing 9.7% and 8.4%, allergic rhinitis symptoms 16.9% and 17.5%, allergic rhinoconjunctivitis symptoms 5.6% and 6.7%, and atopic dermatitis symptoms 5.4% and 3.4%, for...</code> | <code>Leaks of the blood-brain barrier can be detected on postcontrast-enhanced T1-weighted MRIs. Although early disruptions of the blood-brain barrier appear to be an important risk factor for tissue plasminogen activator-related hemorrhages in rodents, little is known about their incidence and consequences in human stroke. This is a retrospective analysis of a prospectively collected stroke database over the past 6 years. In 52 patients, multimodal MRI (including diffusion-weighted, perfusion-weighted, and postcontrast-enhanced T1-weighted MRI to detect blood-brain barrier changes) had been performed immediately before systemic thrombolysis and in 48 patients within a median of 30 minutes (interquartile range: 30 to 60 minutes) after recombinant tissue plasminogen activator treatment. The incidence of symptomatic hemorrhage (SICH), defined as any parenchymal hemorrhage leading to deterioration in the patient's clinical condition, was related to several clinical and imaging variables, inclu...</code> | |
|
| <code>Does synovial fluid glycosaminoglycan concentration correlate with severity of chondropathy or predict progression of osteoarthritis in a canine cruciate deficiency model?</code> | <code>Considerable interest exists today in biochemical or immunochemical tests for monitoring the progression of osteoarthritis (OA). It has been suggested that measurements made on synovial fluid (SF) will more accurately reflect the magnitude of cartilage destruction in an index joint than those performed on serum. However, we have shown that the synovitis that occurs in OA affects the rate of protein clearance from the joint. We tested the hypothesis that if adjusted for clearance rate, the SF concentration of cartilage proteoglycans (PG) estimates severity of chondropathy and predicts progression of cartilage damage more accurately than if clearance is not taken into account. Clearance of radioiodinated serum albumin (RISA), a surrogate for the clearance of PG, was measured in 19 adult dogs at baseline and again 16 weeks and 32 weeks after anterior cruciate ligament transection (ACLT). Severity of chondropathy was determined arthroscopically after 16 weeks of instability and at postmort...</code> | <code>Previous research has demonstrated that chronic cigarette smoking and major depressive disorder (MDD) are each associated with cognitive decrements. Further, these conditions co-occur commonly, though mechanisms in the comorbid condition are poorly understood. There may be distinct, additive, or overlapping factors underlying comorbid cigarette smoking and MDD. The present study investigated the impact of smoking and MDD on executive function and emotion processing. Participants (N=198) were grouped by diagnostic category (MDD and healthy controls, HC) and smoking status (ever-smokers, ES and never-smokers, NS). Participants completed the Facial Emotion Perception Test (FEPT), a measure of emotional processing, and the parametric Go/No-go task (PGNG), a measure of executive function. FEPT performance was analyzed using ANCOVA with accuracy and reaction time as separate dependent variables. Repeated measures MANCOVA was conducted for PGNG with performance measure and task level as depen...</code> | |
|
* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters: |
|
```json |
|
{ |
|
"scale": 20.0, |
|
"similarity_fct": "cos_sim" |
|
} |
|
``` |
|
|
|
### Evaluation Dataset |
|
|
|
#### pub_med_qa_instruction_fotmatted |
|
|
|
* Dataset: [pub_med_qa_instruction_fotmatted](https://huggingface.co/datasets/subhrajit-mohanty/PubMedQA_instruction_fotmatted) at [66d9473](https://huggingface.co/datasets/subhrajit-mohanty/PubMedQA_instruction_fotmatted/tree/66d9473c747f2e73a050b733f8a871cf37281b93) |
|
* Size: 1,000 evaluation samples |
|
* Columns: <code>instruction</code>, <code>context</code>, and <code>context_neg</code> |
|
* Approximate statistics based on the first 1000 samples: |
|
| | instruction | context | context_neg | |
|
|:--------|:-----------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------| |
|
| type | string | string | string | |
|
| details | <ul><li>min: 11 tokens</li><li>mean: 26.49 tokens</li><li>max: 67 tokens</li></ul> | <ul><li>min: 70 tokens</li><li>mean: 317.54 tokens</li><li>max: 512 tokens</li></ul> | <ul><li>min: 45 tokens</li><li>mean: 317.42 tokens</li><li>max: 512 tokens</li></ul> | |
|
* Samples: |
|
| instruction | context | context_neg | |
|
|:------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| |
|
| <code>Is serum 25-hydroxyvitamin D3 related to physical activity and ethnicity but not obesity in a multicultural workforce?</code> | <code>Recent research suggests that body vitamin D levels are decreased in coronary heart disease and diabetes, but it is unclear which cardiovascular risk factors are related to vitamin D status. To examine the relation between vitamin D status and major cardiovascular risk factors. Serum 25-hydroxyvitamin D3, a marker of recent sun exposure and vitamin D status, was measured in 390 New Zealand residents (95 Pacific Islanders, 74 Maori and 221 others mostly of European descent), who were part of a larger cross-sectional survey of a workforce (n = 5677) aged 40-64 years. Serum 25-hydroxyvitamin D3 levels were significantly lower in Pacific Islanders (mean (SE) = 56 (3) nmol/L; p = 0.0001) and Maoris (68 (3) nmol/L; p = 0.036) compared with Europeans (75 (2) nmol/L) after adjusting for age, sex and time of year. Also adjusting for ethnic group, 25-hydroxyvitamin D3 was higher in people doing vigorous (aerobic) leisure physical activities (71 (2) nmol/L; p = 0.0066) and moderate (non-aerobic) ...</code> | <code>Previous follow-up studies indicate that increased visual cortical, ventral cingulate and subcortical responses of depressed individuals to sad facial stimuli, but not happy stimuli could represent reversible markers of disease severity. We hypothesized that greater responses in these areas to sad stimuli, but not happy stimuli, would predict better subsequent clinical outcome. We also explored areas that would predict a poor outcome. Twelve melancholically depressed individuals in the early stages of antidepressant treatment in a secondary care setting participated in two experiments comparing responses to varying intensities of sad and happy facial stimuli, respectively, using event related functional MRI. They repeated the experiments after a mean delay of 12 weeks of treatment. There was a variation in response to treatment. Greater right visual cortex and right subgenual cingulate (R-BA25) responses to sad stimuli, but not happy stimuli, in the early stages of treatment were assoc...</code> | |
|
| <code>Does image subtraction facilitate assessment of volume and density change in ground-glass opacities in chest CT?</code> | <code>To study the impact of image subtraction of registered images on the detection of change in pulmonary ground-glass nodules identified on chest CT. A cohort of 33 individuals (25 men, 8 women; age range 51-75 years) with 37 focal ground-glass opacities (GGO) were recruited from a lung cancer screening trial. For every participant, 1 to 3 follow-up scans were available (total number of pairs, 84). Pairs of scans of the same nodule were registered nonrigidly and then subtracted to enhance differences in size and density. Four observers rated size and density change of the GGO between pairs of scans by visual comparison alone and with additional availability of a subtraction image and indicated their confidence. An independent experienced chest radiologist served as an arbiter having all reader data, clinical data, and follow-up examinations available. Nodule pairs for which the arbiter could not establish definite progression, regression, or stability were excluded from further evaluation...</code> | <code>Betaine serves as a methyl donor in a reaction converting homocysteine to methionine. It is commonly used for the treatment of hyperhomocysteinemia in humans, which indicates it may be associated with reduced risk of atherosclerosis. However, there have been few data regarding its vascular effect. To investigate the effect of betaine supplementation on atherosclerotic lesion in apolipoprotein (apo) E-deficient mice. Four groups of apoE-deficient mice were fed AIN-93G diets supplemented with 0, 1, 2, or 4 g betaine/100 g diet (no, 1, 2, and 4% betaine, respectively). Wild-type C57BL/6 J mice were fed AIN-93G diet (wild-type). Mice were sacrificed after 0, 7, or 14 weeks of the experimental diets. Atherosclerotic lesion area in the aortic sinus, levels of tumor necrosis factor (TNF)-alpha and monocyte chemoattractant protein (MCP)-1 in aorta and serum, serum lipids, and methylation status of TNF-alpha promoter in aorta were determined. Linear regression analysis showed that the higher do...</code> | |
|
| <code>Is overexpression of peptidyl-prolyl isomerase-like 1 associated with the growth of colon cancer cells?</code> | <code>To discover novel therapeutic targets for colon cancers, we previously surveyed expression patterns among 23,000 genes in colon cancer tissues using a cDNA microarray. Among the genes that were up-regulated in the tumors, we selected for this study peptidyl-prolyl isomerase-like 1 (PPIL1) encoding PPIL1, a cyclophilin-related protein. Western blot analysis and immunohistochemical staining using PPIL1-specific antibody showed that PPIL1 protein was frequently overexpressed in colon cancer cells compared with noncancerous epithelial cells of the colon mucosa. Colony formation assay showed a growth-promoting effect of wild-type PPIL1 on NIH3T3 and HEK293 cells. Consistently, transfection of short-interfering RNA specific to PPIL1 into SNUC4 and SNUC5 cells effectively reduced expression of the gene and retarded growth of the colon cancer cells. We further identified two PPIL1-interacting proteins, SNW1/SKIP (SKI-binding protein) and stathmin. SNW1/SKIP is involved in the regulation of tra...</code> | <code>Protocols call for the start of hormonal therapy with levothyroxine after the declaration of brain death. As the hormonal perturbations occur during the process of brain death, the role of the early initiation of levothyroxine therapy (LT) to salvage organs is not well defined. The aim of this study was to evaluate the impact of early LT (before the declaration of brain death) on the number of solid organs procured per donor. We performed an 8-year retrospective analysis of all trauma patients who progressed to brain death. Patients who consented for organ donation, received LT, and donated solid organs were included. Patients were dichotomized into two groups: early LT group, patients who received LT before the declaration of brain death, and late LT group, those who received LT after brain death. The two groups were compared for differences in demographics, clinical characteristics, need for vasopressor, and number of solid organ donation. A total of 100 solid organ donors were ident...</code> | |
|
* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters: |
|
```json |
|
{ |
|
"scale": 20.0, |
|
"similarity_fct": "cos_sim" |
|
} |
|
``` |
|
|
|
### Training Hyperparameters |
|
#### Non-Default Hyperparameters |
|
|
|
- `eval_strategy`: steps |
|
- `per_device_train_batch_size`: 16 |
|
- `per_device_eval_batch_size`: 16 |
|
- `learning_rate`: 2e-05 |
|
- `num_train_epochs`: 1 |
|
- `warmup_ratio`: 0.1 |
|
- `batch_sampler`: no_duplicates |
|
|
|
#### All Hyperparameters |
|
<details><summary>Click to expand</summary> |
|
|
|
- `overwrite_output_dir`: False |
|
- `do_predict`: False |
|
- `eval_strategy`: steps |
|
- `prediction_loss_only`: True |
|
- `per_device_train_batch_size`: 16 |
|
- `per_device_eval_batch_size`: 16 |
|
- `per_gpu_train_batch_size`: None |
|
- `per_gpu_eval_batch_size`: None |
|
- `gradient_accumulation_steps`: 1 |
|
- `eval_accumulation_steps`: None |
|
- `torch_empty_cache_steps`: None |
|
- `learning_rate`: 2e-05 |
|
- `weight_decay`: 0.0 |
|
- `adam_beta1`: 0.9 |
|
- `adam_beta2`: 0.999 |
|
- `adam_epsilon`: 1e-08 |
|
- `max_grad_norm`: 1.0 |
|
- `num_train_epochs`: 1 |
|
- `max_steps`: -1 |
|
- `lr_scheduler_type`: linear |
|
- `lr_scheduler_kwargs`: {} |
|
- `warmup_ratio`: 0.1 |
|
- `warmup_steps`: 0 |
|
- `log_level`: passive |
|
- `log_level_replica`: warning |
|
- `log_on_each_node`: True |
|
- `logging_nan_inf_filter`: True |
|
- `save_safetensors`: True |
|
- `save_on_each_node`: False |
|
- `save_only_model`: False |
|
- `restore_callback_states_from_checkpoint`: False |
|
- `no_cuda`: False |
|
- `use_cpu`: False |
|
- `use_mps_device`: False |
|
- `seed`: 42 |
|
- `data_seed`: None |
|
- `jit_mode_eval`: False |
|
- `use_ipex`: False |
|
- `bf16`: False |
|
- `fp16`: False |
|
- `fp16_opt_level`: O1 |
|
- `half_precision_backend`: auto |
|
- `bf16_full_eval`: False |
|
- `fp16_full_eval`: False |
|
- `tf32`: None |
|
- `local_rank`: 0 |
|
- `ddp_backend`: None |
|
- `tpu_num_cores`: None |
|
- `tpu_metrics_debug`: False |
|
- `debug`: [] |
|
- `dataloader_drop_last`: False |
|
- `dataloader_num_workers`: 0 |
|
- `dataloader_prefetch_factor`: None |
|
- `past_index`: -1 |
|
- `disable_tqdm`: False |
|
- `remove_unused_columns`: True |
|
- `label_names`: None |
|
- `load_best_model_at_end`: False |
|
- `ignore_data_skip`: False |
|
- `fsdp`: [] |
|
- `fsdp_min_num_params`: 0 |
|
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False} |
|
- `fsdp_transformer_layer_cls_to_wrap`: None |
|
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None} |
|
- `deepspeed`: None |
|
- `label_smoothing_factor`: 0.0 |
|
- `optim`: adamw_torch |
|
- `optim_args`: None |
|
- `adafactor`: False |
|
- `group_by_length`: False |
|
- `length_column_name`: length |
|
- `ddp_find_unused_parameters`: None |
|
- `ddp_bucket_cap_mb`: None |
|
- `ddp_broadcast_buffers`: False |
|
- `dataloader_pin_memory`: True |
|
- `dataloader_persistent_workers`: False |
|
- `skip_memory_metrics`: True |
|
- `use_legacy_prediction_loop`: False |
|
- `push_to_hub`: False |
|
- `resume_from_checkpoint`: None |
|
- `hub_model_id`: None |
|
- `hub_strategy`: every_save |
|
- `hub_private_repo`: None |
|
- `hub_always_push`: False |
|
- `gradient_checkpointing`: False |
|
- `gradient_checkpointing_kwargs`: None |
|
- `include_inputs_for_metrics`: False |
|
- `include_for_metrics`: [] |
|
- `eval_do_concat_batches`: True |
|
- `fp16_backend`: auto |
|
- `push_to_hub_model_id`: None |
|
- `push_to_hub_organization`: None |
|
- `mp_parameters`: |
|
- `auto_find_batch_size`: False |
|
- `full_determinism`: False |
|
- `torchdynamo`: None |
|
- `ray_scope`: last |
|
- `ddp_timeout`: 1800 |
|
- `torch_compile`: False |
|
- `torch_compile_backend`: None |
|
- `torch_compile_mode`: None |
|
- `dispatch_batches`: None |
|
- `split_batches`: None |
|
- `include_tokens_per_second`: False |
|
- `include_num_input_tokens_seen`: False |
|
- `neftune_noise_alpha`: None |
|
- `optim_target_modules`: None |
|
- `batch_eval_metrics`: False |
|
- `eval_on_start`: False |
|
- `use_liger_kernel`: False |
|
- `eval_use_gather_object`: False |
|
- `average_tokens_across_devices`: False |
|
- `prompts`: None |
|
- `batch_sampler`: no_duplicates |
|
- `multi_dataset_batch_sampler`: proportional |
|
|
|
</details> |
|
|
|
### Training Logs |
|
| Epoch | Step | Training Loss | Validation Loss | ai-pubmed-validation_cosine_accuracy | |
|
|:-----:|:----:|:-------------:|:---------------:|:------------------------------------:| |
|
| -1 | -1 | - | - | 1.0 | |
|
| 0.2 | 100 | 0.0041 | 0.0031 | 1.0 | |
|
| 0.4 | 200 | 0.0023 | 0.0032 | 1.0 | |
|
| 0.6 | 300 | 0.0027 | 0.0033 | 1.0 | |
|
| 0.8 | 400 | 0.007 | 0.0030 | 1.0 | |
|
| 1.0 | 500 | 0.0024 | 0.0028 | 1.0 | |
|
|
|
|
|
### Framework Versions |
|
- Python: 3.11.11 |
|
- Sentence Transformers: 3.4.1 |
|
- Transformers: 4.48.3 |
|
- PyTorch: 2.5.1+cu124 |
|
- Accelerate: 1.3.0 |
|
- Datasets: 3.3.2 |
|
- Tokenizers: 0.21.0 |
|
|
|
## Citation |
|
|
|
### BibTeX |
|
|
|
#### Sentence Transformers |
|
```bibtex |
|
@inproceedings{reimers-2019-sentence-bert, |
|
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks", |
|
author = "Reimers, Nils and Gurevych, Iryna", |
|
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing", |
|
month = "11", |
|
year = "2019", |
|
publisher = "Association for Computational Linguistics", |
|
url = "https://arxiv.org/abs/1908.10084", |
|
} |
|
``` |
|
|
|
#### MultipleNegativesRankingLoss |
|
```bibtex |
|
@misc{henderson2017efficient, |
|
title={Efficient Natural Language Response Suggestion for Smart Reply}, |
|
author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil}, |
|
year={2017}, |
|
eprint={1705.00652}, |
|
archivePrefix={arXiv}, |
|
primaryClass={cs.CL} |
|
} |
|
``` |
|
|
|
<!-- |
|
## Glossary |
|
|
|
*Clearly define terms in order to be accessible across audiences.* |
|
--> |
|
|
|
<!-- |
|
## Model Card Authors |
|
|
|
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.* |
|
--> |
|
|
|
<!-- |
|
## Model Card Contact |
|
|
|
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.* |
|
--> |